From 20a6cc72917586eb5314be075f96dbb21d16c480 Mon Sep 17 00:00:00 2001 From: "opencode-agent[bot]" Date: Sat, 24 Oct 2026 12:20:37 +0000 Subject: [PATCH] chore(core): refresh bundled models.dev snapshot --- packages/core/src/models-dev/snapshot.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/packages/core/src/models-dev/snapshot.txt b/packages/core/src/models-dev/snapshot.txt index bb7760dec3b..ea0235a7510 100644 --- a/packages/core/src/models-dev/snapshot.txt +++ b/packages/core/src/models-dev/snapshot.txt @@ -1 +1 @@ -{"xai":{"id":"xai","env":["XAI_API_KEY"],"npm":"@ai-sdk/xai","name":"xAI","doc":"https://docs.x.ai/docs/models","models":{"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-4.20-0309-reasoning":{"id":"grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-4.20-multi-agent-0309":{"id":"grok-4.20-multi-agent-0309","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-imagine-video-1.5":{"id":"grok-imagine-video-1.5","type":"video","name":"Grok Imagine Video 1.5","description":"Video model for text-to-video, image-to-video, and reference-to-video generation with optional voice references","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-30","last_updated":"2026-05-30","modalities":{"input":["text","image","audio"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"canonical_model_id":"xai/grok-imagine-video-1.5"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}}},"grok-imagine-image":{"id":"grok-imagine-image","type":"image","name":"Grok Imagine Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"canonical_model_id":"xai/grok-imagine-image"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"grok-4.20-0309-non-reasoning":{"id":"grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-imagine-video-1.5-lite":{"id":"grok-imagine-video-1.5-lite","type":"video","name":"Grok Imagine Video 1.5 Lite","description":"Cost-efficient video model for text-to-video and image-to-video generation","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"canonical_model_id":"xai/grok-imagine-video-1.5-lite"},"grok-imagine-video":{"id":"grok-imagine-video","type":"video","name":"Grok Imagine Video","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0}},"grok-imagine-image-quality":{"id":"grok-imagine-image-quality","type":"image","name":"Grok Imagine Image Quality","description":"Higher-fidelity Grok Imagine image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-03","last_updated":"2026-04-03","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"status":"deprecated","canonical_model_id":"xai/grok-imagine-image-quality"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"}}},"requesty":{"id":"requesty","env":["REQUESTY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://router.requesty.ai/v1","name":"Requesty","doc":"https://requesty.ai/solution/llm-routing/models","models":{"claude-haiku-4-5@eu":{"id":"claude-haiku-4-5@eu","name":"Claude Haiku 4.5 (latest) (EU)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5"},"gpt-4.1-nano@eu":{"id":"gpt-4.1-nano@eu","name":"GPT-4.1 nano (EU)","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.11,"output":0.44,"cache_read":0.0275},"canonical_model_id":"openai/gpt-4.1-nano"},"ling-3.0-tiny":{"id":"ling-3.0-tiny","name":"ling-3.0-tiny","description":"Ling-3.0-tiny is an efficient 7.9B parameter MoE model from inclusionAI with only 1.3B active parameters per token. Built for responsive agents, reliable instruction following and multi turn conversation, with a 256K context window, native function calling, prompt caching and switchable Thinking and Instant modes.","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}},"glm-5.2@eu":{"id":"glm-5.2@eu","name":"GLM-5.2 (EU)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4.2,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"mistral-medium-latest@eu":{"id":"mistral-medium-latest@eu","name":"Mistral Medium (latest) (EU)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.44,"output":2.2,"cache_read":0.44},"canonical_model_id":"mistral/mistral-medium-latest"},"gpt-6-sol@eu":{"id":"gpt-6-sol@eu","name":"GPT-6 Sol (EU)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":12,"cache_read":0.24,"cache_write":3,"tiers":[{"input":4.8,"output":18,"cache_read":0.48,"cache_write":6,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.8,"output":18,"cache_read":0.48,"cache_write":6}},"canonical_model_id":"openai/gpt-6-sol"},"nvidia-nemotron-3-ultra":{"id":"nvidia-nemotron-3-ultra","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":2.5},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gemini-3.7-flash@eu":{"id":"gemini-3.7-flash@eu","name":"Gemini 3.7 Flash (EU)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.825,"output":4.125,"cache_read":0.0825},"canonical_model_id":"google/gemini-3.7-flash"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.1-flash-lite"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.2,"output":4.2,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"kat-coder-pro":{"id":"kat-coder-pro","name":"kat-coder-pro","description":"KAT-Coder-Pro V2 by KwaiKAT is a non-reasoning model optimized for agentic coding. It delivers strong performance on reasoning-style tasks while requiring significantly fewer output tokens than peer models. With the 1210 release, it achieved a score of 64 on the Artificial Analysis Intelligence Index, placing it in the global Top 10 and ranking first among all non-reasoning models.","family":"kat-coder","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-03-27","last_updated":"2026-03-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.3,"output":1.2}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":3,"output":15,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"leanstral-1-5":{"id":"leanstral-1-5","name":"leanstral-1-5","description":"Leanstral 1.5 is an updated Lean 4 formal proof engineering model from Mistral AI, optimized for automated theorem proving and autoformalization. It has 119B total parameters with 6.5B active and supports a 256K token context window. It supports native function calling and structured output.","family":"leanstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}},"claude-fable-5.1@eu":{"id":"claude-fable-5.1@eu","name":"Claude Fable 5.1 (EU)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":0.275,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5-1"},"glm-5.1@eu":{"id":"glm-5.1@eu","name":"GLM-5.1 (EU)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.1"},"nemotron-3-nano-omni@eu":{"id":"nemotron-3-nano-omni@eu","name":"nemotron-3-nano-omni@eu","description":"The most open, efficient, and accurate omni modal reasoning model for agentic AI.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-20","last_updated":"2026-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":300000},"cost":{"input":0.06,"output":0.24,"cache_read":0.06}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"gemini-3.8-flash@eu":{"id":"gemini-3.8-flash@eu","name":"Gemini 3.8 Flash (EU)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.825,"output":4.125,"cache_read":0.0825},"canonical_model_id":"google/gemini-3.8-flash"},"qwen3.8-2.4T-A95B":{"id":"qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"cache_write":1.25,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5}},"canonical_model_id":"xai/grok-4.3"},"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"nvidia-nemotron-3-super-120b-a12b":{"id":"nvidia-nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"claude-sonnet-5@eu":{"id":"claude-sonnet-5@eu","name":"Claude Sonnet 5 (EU)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"qwen3.8-flash-next@eu":{"id":"qwen3.8-flash-next@eu","name":"Qwen3.8 Flash Next (EU)","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"gpt-5.1@eu":{"id":"gpt-5.1@eu","name":"GPT-5.1 (EU)","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":11,"cache_read":0.1375},"canonical_model_id":"openai/gpt-5.1"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"kimi-k2.6@eu":{"id":"kimi-k2.6@eu","name":"Kimi K2.6 (EU)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"cost":{"input":0.95,"output":4,"cache_read":0.95},"canonical_model_id":"moonshotai/kimi-k2.6"},"mistral-medium-latest":{"id":"mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.44,"output":2.2,"cache_read":0.44},"canonical_model_id":"mistral/mistral-medium-latest"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"grok-4.2-beta":{"id":"grok-4.2-beta","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":0.4,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.4,"cache_write":4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"qwen3.5-2b":{"id":"qwen3.5-2b","name":"Qwen3.5 2B","description":"Compact Qwen3.5 multimodal model for fast chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.02,"output":0.1},"canonical_model_id":"alibaba/qwen3.5-2b"},"kimi-k3@eu":{"id":"kimi-k3@eu","name":"Kimi K3 (EU)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":3,"output":15,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"nemotron-3.5-lightning-30b-a3b":{"id":"nemotron-3.5-lightning-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"gemini-2.5-pro@eu":{"id":"gemini-2.5-pro@eu","name":"Gemini 2.5 Pro (EU)","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.25,"output":10,"cache_read":0.31,"cache_write":2.375,"tiers":[{"input":2.5,"output":15,"cache_read":0.62,"cache_write":4.75,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.62,"cache_write":4.75}},"canonical_model_id":"google/gemini-2.5-pro"},"claude-opus-5@eu":{"id":"claude-opus-5@eu","name":"Claude Opus 5 (EU)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"gpt-5.5@eu":{"id":"gpt-5.5@eu","name":"GPT-5.5 (EU)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5.5,"output":33,"cache_read":0.55,"tiers":[{"input":11,"output":49.5,"cache_read":1.1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":11,"output":49.5,"cache_read":1.1}},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"mistral-small-2603@eu":{"id":"mistral-small-2603@eu","name":"Mistral Small 4 (EU)","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.165,"output":0.66,"cache_read":0.165},"canonical_model_id":"mistral/mistral-small-2603"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"laguna-m.1":{"id":"laguna-m.1","name":"Laguna M.1","description":"Poolside's open-weight model for agentic coding and long-horizon work","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-m.1"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0027},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"gpt-5.6-luna@eu":{"id":"gpt-5.6-luna@eu","name":"GPT-5.6 Luna (EU)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"glm-5.3@eu":{"id":"glm-5.3@eu","name":"GLM-5.3 (EU)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.2,"output":4.2,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"devstral-latest@eu":{"id":"devstral-latest@eu","name":"devstral-latest@eu","description":"An enterprise grade text model, that excels at using tools to explore codebases, editing multiple files and power software engineering agents.","family":"devstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.44,"output":2.2,"cache_read":0.44}},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.1},"canonical_model_id":"xai/grok-build-0.1"},"mistral-medium-3-5@eu":{"id":"mistral-medium-3-5@eu","name":"Mistral Medium 3.5 (EU)","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.65,"output":8.25,"cache_read":1.65},"canonical_model_id":"mistral/mistral-medium-3-5"},"ring-2.6-1t":{"id":"ring-2.6-1t","name":"ring-2.6-1t","description":"Inclusion AI ring-2.6-1t","family":"ring","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-08","last_updated":"2026-05-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.5}},"deepseek-v4.1-flash-1017@eu":{"id":"deepseek-v4.1-flash-1017@eu","name":"DeepSeek V4.1 Flash 1017 (EU)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"seed-1.8":{"id":"seed-1.8","name":"Seed 1.8","description":"ByteDance Seed model for multimodal reasoning, long-context analysis, and agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":0.25,"output":2,"cache_read":0.05},"canonical_model_id":"bytedance-seed/seed-1-8"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5@eu":{"id":"gpt-5@eu","name":"GPT-5 (EU)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":11,"cache_read":0.1375},"canonical_model_id":"openai/gpt-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.75,"output":16.5,"cache_read":0.275},"canonical_model_id":"openai/gpt-5.4"},"devstral-latest":{"id":"devstral-latest","name":"devstral-latest","description":"An enterprise grade text model, that excels at using tools to explore codebases, editing multiple files and power software engineering agents.","family":"devstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.44,"output":2.2,"cache_read":0.44}},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"glm-5.2-fast":{"id":"glm-5.2-fast","name":"glm-5.2-fast","description":"GLM-5.2 introduces a robust 1M-token context and advanced, multi-effort coding capabilities to significantly enhance performance on long-horizon tasks. Its new IndexShare architecture and improved MTP layer simultaneously boost efficiency by reducing per-token FLOPs and increasing speculative decoding lengths. A 743B-parameter model in Zhipu AI's GLM series, designed to plan, execute, and iterate autonomously on extended, engineering-grade tasks.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-07-13","last_updated":"2026-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.1,"output":6.6,"cache_read":0.21}},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.5,"output":7,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-5.6-sol@eu":{"id":"gpt-5.6-sol@eu","name":"GPT-5.6 Sol (EU)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5,"tiers":[{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11}},"canonical_model_id":"openai/gpt-5.6-sol"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"claude-sonnet-5-5@eu":{"id":"claude-sonnet-5-5@eu","name":"Claude Sonnet 5.5 (EU)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"inkling-256k":{"id":"inkling-256k","name":"Inkling 256K","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"nemotron-lightning-3.5-30b-a3b":{"id":"nemotron-lightning-3.5-30b-a3b","name":"nemotron-lightning-3.5-30b-a3b","description":"Nemotron-Lightning-3.5-30B-A3B is a 30B-parameter Mixture-of-Experts language model (3B active) from NVIDIA's Nemotron-H family, built on a hybrid Mamba-Transformer architecture for efficient long-context inference. Like other models in the family, it responds to queries by first generating a reasoning trace and then concluding with a final response, with reasoning behavior configurable through a flag in the chat template. It includes a multi-token prediction (MTP) speculative decoding head for low-latency serving.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-15","last_updated":"2026-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.01}},"muse-spark-open":{"id":"muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"claude-opus-4-8@eu":{"id":"claude-opus-4-8@eu","name":"Claude Opus 4.8 (EU)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34,"cache_read":0.07},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"gemini-3.1-flash-lite@eu":{"id":"gemini-3.1-flash-lite@eu","name":"Gemini 3.1 Flash Lite (EU)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.275,"output":1.65,"cache_read":0.0275,"cache_write":0.091663},"canonical_model_id":"google/gemini-3.1-flash-lite"},"deepseek-v4-pro-0813@eu":{"id":"deepseek-v4-pro-0813@eu","name":"DeepSeek V4 Pro 0813 (EU)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":1.75,"output":3.5,"cache_read":0.44},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"mistral-medium-3-5":{"id":"mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.65,"output":8.25,"cache_read":1.65},"canonical_model_id":"mistral/mistral-medium-3-5"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-4o-mini@eu":{"id":"gpt-4o-mini@eu","name":"GPT-4o mini (EU)","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.165,"output":0.66,"cache_read":0.0825},"canonical_model_id":"openai/gpt-4o-mini"},"qwen3.8-2.4T-A95B@eu":{"id":"qwen3.8-2.4T-A95B@eu","name":"Qwen3.8 2.4T A95B (EU)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":2.5,"output":6,"cache_read":0.63},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"gpt-6-luna@eu":{"id":"gpt-6-luna@eu","name":"GPT-6 Luna (EU)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.12,"output":0.6,"cache_read":0.012,"cache_write":0.15,"tiers":[{"input":0.24,"output":0.9,"cache_read":0.024,"cache_write":0.3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.24,"output":0.9,"cache_read":0.024,"cache_write":0.3}},"canonical_model_id":"openai/gpt-6-luna"},"nemotron-3-ultra-550b-a55b":{"id":"nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":20480},"cost":{"input":0,"output":0},"canonical_model_id":"meta/muse-glimmer-30b"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"claude-sonnet-4@eu":{"id":"claude-sonnet-4@eu","name":"Claude Sonnet 4 (latest) (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-5.4@eu":{"id":"gpt-5.4@eu","name":"GPT-5.4 (EU)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.75,"output":16.5,"cache_read":0.275,"tiers":[{"input":5.5,"output":24.75,"cache_read":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5.5,"output":24.75,"cache_read":0.55}},"canonical_model_id":"openai/gpt-5.4"},"o4-mini@eu":{"id":"o4-mini@eu","name":"o4-mini (EU)","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.21,"output":4.84,"cache_read":0.3025},"canonical_model_id":"openai/o4-mini"},"glm-5.3-flash@eu":{"id":"glm-5.3-flash@eu","name":"GLM-5.3-Flash (EU)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0.2,"output":0.6,"cache_read":0.07},"canonical_model_id":"zhipuai/glm-5.3-flash"},"mistral-large-4@eu":{"id":"mistral-large-4@eu","name":"Mistral Large 4 (EU)","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"gemini-2.5-flash@eu":{"id":"gemini-2.5-flash@eu","name":"Gemini 2.5 Flash (EU)","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.075,"cache_write":0.55},"canonical_model_id":"google/gemini-2.5-flash"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"ling-2.6-flash":{"id":"ling-2.6-flash","name":"ling-2.6-flash","description":"Inclusion AI ling-2.6-flash","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3}},"gemini-4-argon@eu":{"id":"gemini-4-argon@eu","name":"Gemini 4 Argon (EU)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2.2,"output":11,"cache_read":0.11},"canonical_model_id":"google/gemini-4-argon"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"cache_read":30},"canonical_model_id":"openai/gpt-5.4-pro"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"seed-2.0-mini":{"id":"seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"nemotron-3-ultra-nvfp4":{"id":"nemotron-3-ultra-nvfp4","name":"nemotron-3-ultra-nvfp4","description":"Nemotron-3-Ultra-550B-A55B-NVFP4 is a frontier-scale large language model (LLM) trained by NVIDIA, designed to deliver strong agentic, reasoning, and conversational capabilities. It is optimized for the most demanding workloads, including complex multi-step agents, long-context analysis, and high-accuracy reasoning over code, math, and science. The model employs a hybrid Latent Mixture-of-Experts (LatentMoE) architecture, utilizing interleaved Mamba-2 and MoE layers, along with select Attention layers. Like the Super model, the Ultra model incorporates Multi-Token Prediction (MTP) layers for faster text generation and improved quality, and it is trained using an NVFP4 pre-training recipe to maximize compute efficiency. The model has 55B active parameters and 550B parameters in total.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":2.4,"cache_read":0.12}},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":4.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":9,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":9}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2},"canonical_model_id":"xai/grok-4.7"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"nemotron-3-nano-omni-30b-a3b-reasoning":{"id":"nemotron-3-nano-omni-30b-a3b-reasoning","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":20480},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"fugu-ultra":{"id":"fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"nemotron-3-nano-omni":{"id":"nemotron-3-nano-omni","name":"nemotron-3-nano-omni","description":"The most open, efficient, and accurate omni modal reasoning model for agentic AI.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-20","last_updated":"2026-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":300000},"cost":{"input":0.06,"output":0.24,"cache_read":0.06}},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"nemotron-3.5-content-safety":{"id":"nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"laguna-xs.2":{"id":"laguna-xs.2","name":"Laguna XS.2","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs.2"},"seed-2.0-pro":{"id":"seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"claude-opus-4-5@eu":{"id":"claude-opus-4-5@eu","name":"Claude Opus 4.5 (latest) (EU)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-5"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":2.55,"cache_read":0.16},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gpt-5.6-terra@eu":{"id":"gpt-5.6-terra@eu","name":"GPT-5.6 Terra (EU)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3.8-flash-next":{"id":"qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0.2,"output":0.6,"cache_read":0.07},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gpt-6.1-sol@eu":{"id":"gpt-6.1-sol@eu","name":"GPT-6.1 Sol (EU)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":12,"cache_read":0.12,"cache_write":3,"tiers":[{"input":4.8,"output":18,"cache_read":0.24,"cache_write":6,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.8,"output":18,"cache_read":0.24,"cache_write":6}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.32,"output":1.28,"cache_read":0.032,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.7-plus"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":4.5},"canonical_model_id":"google/gemini-3-pro-image"},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":32768},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"leanstral-1-5@eu":{"id":"leanstral-1-5@eu","name":"leanstral-1-5@eu","description":"Leanstral 1.5 is an updated Lean 4 formal proof engineering model from Mistral AI, optimized for automated theorem proving and autoformalization. It has 119B total parameters with 6.5B active and supports a 256K token context window. It supports native function calling and structured output.","family":"leanstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":2},"canonical_model_id":"google/gemini-3.1-flash-image"},"kimi-k2.7-code@eu":{"id":"kimi-k2.7-code@eu","name":"Kimi K2.7 Code (EU)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.25,"output":4.5,"cache_read":0.31},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4.1-flash@eu":{"id":"deepseek-v4.1-flash@eu","name":"DeepSeek V4.1 Flash (EU)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"claude-opus-5-5@eu":{"id":"claude-opus-5-5@eu","name":"Claude Opus 5.5 (EU)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-sonnet-4-5@eu":{"id":"claude-sonnet-4-5@eu","name":"Claude Sonnet 4.5 (latest) (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.3,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.6,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.6,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.165,"output":0.66,"cache_read":0.165},"canonical_model_id":"mistral/mistral-small-2603"},"minimax-m3@eu":{"id":"minimax-m3@eu","name":"MiniMax-M3 (EU)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.4,"output":2,"cache_read":0.1},"canonical_model_id":"minimax/MiniMax-M3"},"gpt-5-nano@eu":{"id":"gpt-5-nano@eu","name":"GPT-5 Nano (EU)","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":0.055,"output":0.44,"cache_read":0.0055},"canonical_model_id":"openai/gpt-5-nano"},"gpt-4.1-mini@eu":{"id":"gpt-4.1-mini@eu","name":"GPT-4.1 mini (EU)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.44,"output":1.76,"cache_read":0.11},"canonical_model_id":"openai/gpt-4.1-mini"},"nemotron-3-super-120b-a12b":{"id":"nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"gemini-2.5-flash-lite@eu":{"id":"gemini-2.5-flash-lite@eu","name":"Gemini 2.5 Flash-Lite (EU)","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.18333},"canonical_model_id":"google/gemini-2.5-flash-lite"},"gpt-4.1@eu":{"id":"gpt-4.1@eu","name":"GPT-4.1 (EU)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2.2,"output":8.8,"cache_read":0.55},"canonical_model_id":"openai/gpt-4.1"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"minimax-m2.7-highspeed":{"id":"minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":1.2},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5.5,"output":33,"cache_read":0.55},"canonical_model_id":"openai/gpt-5.5"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.13,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.63,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-fable-5.1":{"id":"claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26,"output":2.6},"canonical_model_id":"alibaba/qwen3.5-27b"},"deepseek-v4-flash-0731@eu":{"id":"deepseek-v4-flash-0731@eu","name":"DeepSeek V4 Flash 0731 (EU)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-3.5-flash-lite@eu":{"id":"gemini-3.5-flash-lite@eu","name":"Gemini 3.5 Flash Lite (EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.033},"canonical_model_id":"google/gemini-3.5-flash-lite"},"kimi-k2.5@eu":{"id":"kimi-k2.5@eu","name":"Kimi K2.5 (EU)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.72,"output":3.6,"cache_read":0.72,"cache_write":0.72},"canonical_model_id":"moonshotai/kimi-k2.5"},"gpt-5-mini@eu":{"id":"gpt-5-mini@eu","name":"GPT-5 Mini (EU)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":0.275,"output":2.2,"cache_read":0.0275},"canonical_model_id":"openai/gpt-5-mini"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"ling-2.6-1t":{"id":"ling-2.6-1t","name":"ling-2.6-1t","description":"Inclusion AI ling-2.6-1t","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.5}},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":1.583},"canonical_model_id":"google/gemini-3.5-flash"},"claude-haiku-5-5@eu":{"id":"claude-haiku-5-5@eu","name":"Claude Haiku 5.5 (EU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.02,"cache_write":0.14,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.06,"cache_write":0.69,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gemini-3.5-flash@eu":{"id":"gemini-3.5-flash@eu","name":"Gemini 3.5 Flash (EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.65,"output":9.9,"cache_read":0.165,"cache_write":1.7413},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"claude-opus-4-7@eu":{"id":"claude-opus-4-7@eu","name":"Claude Opus 4.7 (EU)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"claude-sonnet-4-6@eu":{"id":"claude-sonnet-4-6@eu","name":"Claude Sonnet 4.6 (EU)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.3,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"deepseek-v4-pro@eu":{"id":"deepseek-v4-pro@eu","name":"DeepSeek V4 Pro (EU)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.75,"output":3.5,"cache_read":0.44},"canonical_model_id":"deepseek/deepseek-v4-pro"},"claude-fable-5@eu":{"id":"claude-fable-5@eu","name":"Claude Fable 5 (EU)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.5"},"claude-opus-4-6@eu":{"id":"claude-opus-4-6@eu","name":"Claude Opus 4.6 (EU)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"seed-2.0-code":{"id":"seed-2.0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"synthetic":{"id":"synthetic","env":["SYNTHETIC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.synthetic.new/openai/v1","name":"Synthetic","doc":"https://synthetic.new/pricing","models":{"hf:moonshotai/Kimi-K3":{"id":"hf:moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":3,"output":15,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"hf:moonshotai/Kimi-K2.7-Code":{"id":"hf:moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.95},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"hf:openai/gpt-oss-120b":{"id":"hf:openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.1,"output":0.1,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-120b"},"hf:Qwen/Qwen3.6-27B":{"id":"hf:Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.45,"output":3.6,"cache_read":0.45},"canonical_model_id":"alibaba/qwen3.6-27b"},"hf:MiniMaxAI/MiniMax-M3":{"id":"hf:MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"status":"deprecated","cost":{"input":0.6,"output":1.2,"cache_read":0.6},"canonical_model_id":"minimax/MiniMax-M3"},"hf:zai-org/GLM-5.3":{"id":"hf:zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"hf:zai-org/GLM-4.7-Flash":{"id":"hf:zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":65536},"cost":{"input":0.1,"output":0.5,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-4.7-flash"},"hf:zai-org/GLM-5.3-Flash":{"id":"hf:zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.15,"output":0.5,"cache_read":0.04},"canonical_model_id":"zhipuai/glm-5.3-flash"},"hf:zai-org/GLM-5.2":{"id":"hf:zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.2"},"hf:nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4":{"id":"hf:nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":1,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"hf:deepseek-ai/DeepSeek-V4.1-Flash":{"id":"hf:deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.6,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"hf:deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"hf:deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.6,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"}}},"lmstudio":{"id":"lmstudio","env":["LMSTUDIO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://127.0.0.1:1234/v1","name":"LMStudio","doc":"https://lmstudio.ai/models","models":{"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-20b"},"qwen/qwen3-30b-a3b-2507":{"id":"qwen/qwen3-30b-a3b-2507","name":"Qwen3 30B A3B 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3-coder-30b":{"id":"qwen/qwen3-coder-30b","name":"Qwen3 Coder 30B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"}}},"wafer.ai":{"id":"wafer.ai","env":["WAFER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://pass.wafer.ai/v1","name":"Wafer","doc":"https://docs.wafer.ai/wafer-pass","models":{"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Kimi K2.6 sparse MoE model with a 262K context window. Available serverless and not included in standard Wafer Pass. Non-ZDR only: requests with `Wafer-ZDR: required` are rejected.","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.14,"output":4.8,"cache_read":0.19,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"General Language Model 5.1 — high-quality bilingual (EN/ZH) generation with strong coding and reasoning capabilities.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1,"output":3.2,"cache_read":0.1,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"status":"deprecated","cost":{"input":0.33,"output":1.32,"cache_read":0.07,"cache_write":0,"tiers":[{"input":0.66,"output":2.64,"cache_read":0.13,"cache_write":0,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.66,"output":2.64,"cache_read":0.13,"cache_write":0}},"canonical_model_id":"minimax/MiniMax-M3"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm5.2-fast":{"id":"glm5.2-fast","name":"GLM5.2-Fast","description":"The same model served for high TPS.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":3,"output":10.25,"cache_read":0.5,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"kuae-cloud-coding-plan":{"id":"kuae-cloud-coding-plan","env":["KUAE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://coding-plan-endpoint.kuaecloud.net/v1","name":"KUAE Cloud Coding Plan","doc":"https://docs.mthreads.com/kuaecloud/kuaecloud-doc-online/coding_plan/","models":{"GLM-4.7":{"id":"GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"}}},"agentrouter":{"id":"agentrouter","env":["AGENTROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://agentrouter.org/v1","name":"AgentRouter","doc":"https://agentrouter.org/docs/opencode.html","models":{"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-flash"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://agentrouter.org/v1"},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://agentrouter.org/v1"},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-5.6-sol"}}},"zenmux":{"id":"zenmux","env":["ZENMUX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://zenmux.ai/api/v1","name":"ZenMux","doc":"https://docs.zenmux.ai","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.075,"output":0.3,"cache_read":0.0015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek-V3.2-Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163000,"output":64000},"status":"deprecated","cost":{"input":0.22,"output":0.33}},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":0.293,"output":0.4395,"cache_read":0.0293},"canonical_model_id":"deepseek/deepseek-v3.2"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.2-fast-non-reasoning":{"id":"x-ai/grok-4.2-fast-non-reasoning","name":"Grok 4.2 Fast Non Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":4,"output":12,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"x-ai/grok-imagine-image-2.0":{"id":"x-ai/grok-imagine-image-2.0","type":"image","name":"Grok Imagine Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":66000,"output":0},"canonical_model_id":"xai/grok-imagine-image-2.0"},"x-ai/grok-voice-tts-1.0":{"id":"x-ai/grok-voice-tts-1.0","type":"speech","name":"Grok Voice TTS 1.0","description":"Convert text into spoken audio with a single API call. The API supports a rich set of expressive voices, inline speech tags for fine-grained delivery control, and output formats from high-fidelity MP3 to telephony-optimized μ-law.","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":15000,"output":15000},"canonical_model_id":"xai/grok-tts"},"x-ai/grok-voice-stt-1.0":{"id":"x-ai/grok-voice-stt-1.0","type":"transcription","name":"Grok Voice STT 1.0","description":"Speech-to-text model with word-level timestamps, speaker diarization, multichannel audio, and streaming","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":15000,"output":15000},"canonical_model_id":"xai/grok-voice-transcribe-2.0"},"x-ai/grok-4.2-fast":{"id":"x-ai/grok-4.2-fast","name":"Grok 4.2 Fast","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":4,"output":12,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.5"},"inclusionai/ring-2.6-1t":{"id":"inclusionai/ring-2.6-1t","name":"inclusionAI: Ring-2.6-1T","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-12-31","release_date":"2026-05-07","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65000},"cost":{"input":0.1318155,"output":1.0984625,"cache_read":0.0263631}},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.5-lightning":{"id":"minimax/minimax-m2.5-lightning","name":"MiniMax M2.5 highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131070},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":512000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131070},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.611,"output":2.4439,"cache_read":0.0611,"cache_write":0.3819},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"volcengine/doubao-seed-2.0-pro":{"id":"volcengine/doubao-seed-2.0-pro","name":"Doubao-Seed-2.0-pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02-14","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.45,"output":2.24,"cache_read":0.09,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"volcengine/doubao-seed-1.8":{"id":"volcengine/doubao-seed-1.8","name":"Doubao-Seed-1.8","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.11,"output":0.28,"cache_read":0.02,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-1-8"},"volcengine/doubao-seed-2.0-lite":{"id":"volcengine/doubao-seed-2.0-lite","name":"Doubao-Seed-2.0-lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02-14","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.09,"output":0.51,"cache_read":0.02,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"volcengine/doubao-seed-code":{"id":"volcengine/doubao-seed-code","name":"Doubao-Seed-Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-11-11","last_updated":"2025-11-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.17,"output":1.12,"cache_read":0.03}},"volcengine/doubao-seed-2.0-code":{"id":"volcengine/doubao-seed-2.0-code","name":"Doubao Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.9,"output":4.48},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"volcengine/doubao-seed-2.0-mini":{"id":"volcengine/doubao-seed-2.0-mini","name":"Doubao-Seed-2.0-mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02-14","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.03,"output":0.28,"cache_read":0.01,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"baidu/ernie-5.0-thinking-preview":{"id":"baidu/ernie-5.0-thinking-preview","name":"ERNIE 5.0","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-01-22","last_updated":"2026-01-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":0.84,"output":3.37,"tiers":[{"input":1.41,"output":5.62,"tier":{"type":"context","size":32000}}]}},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":2,"cache_read":0.08,"tiers":[{"input":0.8,"output":4,"cache_read":0.16,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.8,"output":4,"cache_read":0.16}},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1,"output":3,"cache_read":0.2,"tiers":[{"input":2,"output":6,"cache_read":0.4,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.4}},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":1.25,"output":10,"cache_read":0.13,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":1,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1-Codex-Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2-Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/chat-latest":{"id":"openai/chat-latest","name":"Chat Latest (GPT-5.5 Instant)","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-05-05","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5-instant"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"bytedance/doubao-seed-2.0-pro":{"id":"bytedance/doubao-seed-2.0-pro","name":"Doubao-Seed-2.0-pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.45,"output":2.24,"cache_read":0.09,"tiers":[{"input":0.67,"output":3.36,"cache_read":0.14,"tier":{"type":"context","size":32000}},{"input":1.34,"output":6.71,"cache_read":0.27,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"bytedance/doubao-seed-1.8":{"id":"bytedance/doubao-seed-1.8","name":"Doubao-Seed-1.8","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.11,"output":0.28,"cache_read":0.023,"tiers":[{"input":0.17,"output":2.27,"cache_read":0.023,"tier":{"type":"context","size":32000}},{"input":0.34,"output":3.41,"cache_read":0.023,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-8"},"bytedance/doubao-seed-2.0-lite":{"id":"bytedance/doubao-seed-2.0-lite","name":"Doubao-Seed-2.0-lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.09,"output":0.51,"cache_read":0.02,"tiers":[{"input":0.13,"output":0.76,"cache_read":0.03,"tier":{"type":"context","size":32000}},{"input":0.25,"output":1.51,"cache_read":0.05,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance/doubao-seed-code":{"id":"bytedance/doubao-seed-code","name":"Doubao-Seed-Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-11-11","last_updated":"2025-11-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.17,"output":1.12,"cache_read":0.034,"tiers":[{"input":0.2,"output":1.69,"cache_read":0.034,"tier":{"type":"context","size":32000}},{"input":0.39,"output":2.25,"cache_read":0.034,"tier":{"type":"context","size":128000}}]}},"bytedance/doubao-seed-2.0-code":{"id":"bytedance/doubao-seed-2.0-code","name":"Doubao Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.45,"output":2.24,"cache_read":0.09,"tiers":[{"input":0.67,"output":3.36,"cache_read":0.14,"tier":{"type":"context","size":32000}},{"input":1.34,"output":6.71,"cache_read":0.27,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"bytedance/doubao-seed-2.0-mini":{"id":"bytedance/doubao-seed-2.0-mini","name":"Doubao-Seed-2.0-mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.03,"output":0.28,"cache_read":0.01,"tiers":[{"input":0.06,"output":0.56,"cache_read":0.02,"tier":{"type":"context","size":32000}},{"input":0.12,"output":1.12,"cache_read":0.03,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262140,"output":262140},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code HighSpeed","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6-Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.4,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.4,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.25,"output":3.75,"cache_read":0.25,"cache_write":1.5625},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3-Coder-Plus","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3-Max-Thinking","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":2.4,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625}},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1020000,"output":1020000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.58,"output":2.6,"cache_read":0.14,"tiers":[{"input":0.87,"output":3.18,"cache_read":0.22,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.73,"output":3.19,"cache_read":0.174,"tiers":[{"input":1.02,"output":3.77,"cache_read":0.261,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM 5.3 FlashX","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.375,"output":1.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.8781,"output":3.5126,"cache_read":0.1903,"tiers":[{"input":1.1709,"output":4.098,"cache_read":0.2927,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.7-flashx":{"id":"z-ai/glm-4.7-flashx","name":"GLM 4.7 FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.0728,"output":0.4367,"cache_read":0.0146},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.2911,"output":1.1645,"cache_read":0.0582,"tiers":[{"input":0.5823,"output":2.3291,"cache_read":0.1165,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.2911,"output":1.1645,"cache_read":0.0582,"tiers":[{"input":0.5823,"output":2.3291,"cache_read":0.1165,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-4.7-flash-free":{"id":"z-ai/glm-4.7-flash-free","name":"GLM 4.7 Flash (Free)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-image":{"id":"z-ai/glm-image","type":"image","name":"GLM-Image","description":"GLM-Image is an image generation model adopts a hybrid autoregressive + diffusion decoder architecture. In general image generation quality, GLM‑Image aligns with mainstream latent diffusion approaches, but it shows significant advantages in text-rendering and knowledge‑intensive generation scenarios. It performs especially well in tasks requiring precise semantic understanding and complex information expression, while maintaining strong capabilities in high‑fidelity and fine‑grained detail generation. In addition to text‑to‑image generation, GLM‑Image also supports a rich set of image‑to‑image tasks including image editing, style transfer, identity‑preserving generation, and multi‑subject consistency.","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-14","last_updated":"2026-01-14","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":10240,"output":0},"provider":{"api":"https://zenmux.ai/api/vertex-ai"},"canonical_model_id":"zhipuai/glm-image"},"z-ai/glm-4.6v-flash":{"id":"z-ai/glm-4.6v-flash","name":"GLM 4.6V FlashX","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.0218,"output":0.2184,"cache_read":0.0044,"tiers":[{"input":0.0437,"output":0.4367,"cache_read":0.0044,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.075,"output":0.25,"cache_read":0.015},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.726,"output":3.1946,"cache_read":0.1743,"tiers":[{"input":1.0165,"output":3.7754,"cache_read":0.2614,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.2911,"output":1.1645,"cache_read":0.0582,"tiers":[{"input":0.5823,"output":2.3291,"cache_read":0.1165,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM 4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.1456,"output":0.4367,"cache_read":0.0291,"tiers":[{"input":0.2911,"output":0.8734,"cache_read":0.0582,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.1165,"output":0.2911,"cache_read":0.0233,"tiers":[{"input":0.1747,"output":1.1645,"cache_read":0.0349,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-4.6v-flash-free":{"id":"z-ai/glm-4.6v-flash-free","name":"GLM 4.6V Flash (Free)","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"tiers":[{"input":0,"output":0,"cache_read":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v-flash"}}},"umans-ai-coding-plan":{"id":"umans-ai-coding-plan","env":["UMANS_AI_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.code.umans.ai/v1","name":"Umans AI Coding Plan","doc":"https://app.umans.ai/offers/code/docs","models":{"umans-kimi-k3":{"id":"umans-kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"umans-coder":{"id":"umans-coder","name":"Umans Coder","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-glm-5.3-flash":{"id":"umans-glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-qwen3.6-35b-a3b":{"id":"umans-qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"umans-glm-5.3":{"id":"umans-glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"umans-deepseek-v4-flash-0731":{"id":"umans-deepseek-v4-flash-0731","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"umans-deepseek-v4.1-flash-1017":{"id":"umans-deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"umans-flash":{"id":"umans-flash","name":"Umans Flash","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"umans-deepseek-v4.1-flash":{"id":"umans-deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"}}},"deepseek":{"id":"deepseek","env":["DEEPSEEK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.deepseek.com","name":"DeepSeek","doc":"https://api-docs.deepseek.com/quick_start/pricing","models":{"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.66,"output":1.98,"reasoning":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"status":"deprecated","cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"status":"deprecated","cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-flash":{"id":"deepseek-flash","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"}}},"baseten":{"id":"baseten","env":["BASETEN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.baseten.co/v1","name":"Baseten","doc":"https://docs.baseten.co/inference/model-apis/overview","models":{"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"OpenAI GPT 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128072,"output":128072},"cost":{"input":0.1,"output":0.5,"cache_read":0.1},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.32,"output":3.96,"cache_read":0.132},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.3,"output":1.2,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4.1-Flash-1017-Fast":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017-Fast","name":"DeepSeek V4.1 Flash 1017 Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-V4.1-Flash-Fast":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-Fast","name":"DeepSeek V4.1 Flash Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B","name":"Nemotron Ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":202800},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM 5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.2-Fast":{"id":"zai-org/GLM-5.2-Fast","name":"GLM 5.2 Fast","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5.3-Fast":{"id":"zai-org/GLM-5.3-Fast","name":"GLM 5.3 Fast","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.3"}}},"azure":{"id":"azure","env":["AZURE_RESOURCE_NAME","AZURE_API_KEY"],"npm":"@ai-sdk/azure","name":"Azure","doc":"https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/models","models":{"deepseek-v3.2-speciale":{"id":"deepseek-v3.2-speciale","name":"DeepSeek-V3.2-Speciale","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek-V4.1-Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.71,"output":0.71}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-3.5-turbo-1106":{"id":"gpt-3.5-turbo-1106","name":"GPT-3.5 Turbo 1106","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":1,"output":2}},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-01-14","last_updated":"2026-01-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"cohere-embed-v3-english":{"id":"cohere-embed-v3-english","type":"embedding","name":"Embed v3 English","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek-V4-Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":1.74,"output":3.48},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gpt-image-2.5-flare":{"id":"gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"cohere-embed-v-4-0":{"id":"cohere-embed-v-4-0","type":"embedding","name":"Embed v4","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":1536},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.12,"output":0}},"phi-4-reasoning-plus":{"id":"phi-4-reasoning-plus","name":"Phi-4-reasoning-plus","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25}},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-08-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"gpt-3.5-turbo-0125":{"id":"gpt-3.5-turbo-0125","name":"GPT-3.5 Turbo 0125","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":0.5,"output":1.5}},"model-router":{"id":"model-router","name":"Model Router","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"model-router","attachment":true,"reasoning":false,"tool_call":true,"release_date":"2025-05-19","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":16384},"provider":{"npm":"@ai-sdk/azure","shape":"completions"},"cost":{"input":0.14,"output":0}},"gpt-image-2.5-sunburst":{"id":"gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"phi-4-mini":{"id":"phi-4-mini","name":"Phi-4-mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"phi-4-mini-reasoning":{"id":"phi-4-mini-reasoning","name":"Phi-4-mini-reasoning","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"llama-4-maverick-17b-128e-instruct-fp8":{"id":"llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B 128E Instruct FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":16384},"provider":{"shape":"completions"},"cost":{"input":0.25,"output":1}},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-chat-latest":{"id":"gpt-chat-latest","name":"GPT Chat Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-05-05","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"beta","cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5-instant"},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.02,"output":0}},"llama-4-scout-17b-16e-instruct":{"id":"llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B 16E Instruct","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.78}},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek-V4-Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.19,"output":0.51},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"mistral-medium-2505":{"id":"mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.4,"output":2}},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-image-1":{"id":"gpt-image-1","type":"image","name":"GPT-Image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":40,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1"},"claude-mythos-5":{"id":"claude-mythos-5","name":"Claude Mythos 5","description":"Restricted Claude model for advanced cybersecurity and biology research workflows","family":"claude-mythos","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-mythos-5"},"gpt-3.5-turbo-instruct":{"id":"gpt-3.5-turbo-instruct","name":"GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-09-21","last_updated":"2023-09-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":4096},"status":"deprecated","cost":{"input":1.5,"output":2}},"phi-4-reasoning":{"id":"phi-4-reasoning","name":"Phi-4-reasoning","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"gpt-4-turbo":{"id":"gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"codestral-2501":{"id":"codestral-2501","name":"Codestral 25.01","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"provider":{"shape":"completions"},"cost":{"input":0.3,"output":0.9}},"cohere-command-a":{"id":"cohere-command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"provider":{"shape":"completions"},"cost":{"input":2.5,"output":10}},"ministral-3b":{"id":"ministral-3b","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.04,"output":0.04}},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"phi-4-multimodal":{"id":"phi-4-multimodal","name":"Phi-4-multimodal","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"phi","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.08,"output":0.32,"input_audio":4}},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}},"phi-4":{"id":"phi-4","name":"Phi-4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"status":"beta","provider":{"shape":"completions"},"canonical_model_id":"xai/grok-4.7"},"o1":{"id":"o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5}},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-07-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}}},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-4-turbo-vision":{"id":"gpt-4-turbo-vision","name":"GPT-4 Turbo Vision","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"cohere-embed-v3-multilingual":{"id":"cohere-embed-v3-multilingual","type":"embedding","name":"Embed v3 Multilingual","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"mistral-small-2503":{"id":"mistral-small-2503","name":"Mistral Small 3.1","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-03-01","last_updated":"2025-03-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.1,"output":0.3}},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-31","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"grok-4-20-reasoning":{"id":"grok-4-20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"codex-mini":{"id":"codex-mini","name":"Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-05-16","last_updated":"2025-05-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.5,"output":6,"cache_read":0.375}},"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"GPT-Image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":32,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1.5"},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"cost":{"input":0.13,"output":0}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005}},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"grok-4-20-non-reasoning":{"id":"grok-4-20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"text-embedding-ada-002":{"id":"text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.1,"output":0}},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"bothub":{"id":"bothub","env":["BOTHUB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://openai.bothub.ru/v1","name":"Bothub","doc":"https://bothub.ru/models","models":{"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4243,"output":4.4765},"canonical_model_id":"zhipuai/glm-5.3"},"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"meta/muse-spark-1.3"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2677,"output":1.6064},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.7671,"output":5.3011},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.589,"output":1.7671},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1566,"output":0.522},"canonical_model_id":"zhipuai/glm-5.3-flash"},"nemotron-3-ultra-550b-a55b:free":{"id":"nemotron-3-ultra-550b-a55b:free","name":"Nemotron 3 Ultra (free)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gemma-4-31b-it:free":{"id":"gemma-4-31b-it:free","name":"Gemma 4 31B IT (free)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"}}},"sakana":{"id":"sakana","env":["SAKANA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.sakana.ai/v1","name":"Sakana AI","doc":"https://console.sakana.ai/models","models":{"fugu":{"id":"fugu","name":"Fugu","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"provider":{"shape":"responses"},"canonical_model_id":"sakana/fugu"},"sakana-namazu":{"id":"sakana-namazu","name":"Sakana Namazu","description":"Japanese-specialized reasoning model based on Kimi K2.6 and tuned for Japanese language, culture, and business workflows","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"fugu-ultra-20260615":{"id":"fugu-ultra-20260615","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"provider":{"shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"fugu-ultra":{"id":"fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"provider":{"shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"}}},"nano-gpt":{"id":"nano-gpt","env":["NANO_GPT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://nano-gpt.com/api/v1","name":"NanoGPT","doc":"https://docs.nano-gpt.com","models":{"gemini-2.5-flash-nothinking":{"id":"gemini-2.5-flash-nothinking","name":"Gemini 2.5 Flash (No Thinking)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"doubao-1.5-pro-32k":{"id":"doubao-1.5-pro-32k","name":"Doubao 1.5 Pro 32k","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-11-20","last_updated":"2025-01-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.1343,"output":0.3349,"cache_read":0.06715}},"mistral-code-agent-latest":{"id":"mistral-code-agent-latest","name":"Mistral Code Agent Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.4,"output":2,"cache_read":0.2}},"gemini-2.5-flash-lite-preview-06-17":{"id":"gemini-2.5-flash-lite-preview-06-17","name":"Gemini 2.5 Flash Lite Preview","description":"First preview of Gemini 2.5 Flash-Lite, the low-cost, low-latency 2.5 model","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-06-17"},"Gemma-4-26B-A4B-MeroMero:thinking":{"id":"Gemma-4-26B-A4B-MeroMero:thinking","name":"Gemma 4 26B A4B MeroMero Thinking","description":"Gemma 4 26B A4B MeroMero with thinking enabled for more deliberate emotive dialogue, relationship scenes, creative writing, and multimodal roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"glm-4-air-0111":{"id":"glm-4-air-0111","name":"GLM 4 Air 0111","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-01-11","last_updated":"2025-01-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":0.1394,"output":0.1394,"cache_read":0.0697}},"gemma-4-12b-it-station-keeper":{"id":"gemma-4-12b-it-station-keeper","name":"Gemma 4 12B StationKeeper","description":"Gemma 4 12B StationKeeper is an open-weight roleplay finetune with image understanding, tool calling, optional reasoning, and a 131,072-token context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.025}},"nano-gpt-help":{"id":"nano-gpt-help","name":"NanoGPT Help","description":"Text-only NanoGPT support assistant. Questions are processed by the Help inference provider; do not paste secrets or account credentials. Covers the website, models, API, pricing, memory, media generation, and support.","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-06-06","last_updated":"2026-06-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":6000,"input":6000,"output":512},"cost":{"input":0,"output":0}},"doubao-1.5-vision-pro-32k":{"id":"doubao-1.5-vision-pro-32k","name":"Doubao 1.5 Vision Pro 32k","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-11-20","last_updated":"2025-01-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.459,"output":1.377,"cache_read":0.2295}},"venice-uncensored":{"id":"venice-uncensored","name":"Venice Uncensored","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"venice","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-10-01","last_updated":"2025-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.4,"output":1.8,"cache_read":0.4}},"gemini-2.5-flash-lite-preview-09-2025":{"id":"gemini-2.5-flash-lite-preview-09-2025","name":"Gemini 2.5 Flash Lite Preview (09/2025)","description":"September 2025 preview of Gemini 2.5 Flash-Lite with better instruction following and multimodal quality","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-09-2025"},"gemma-4-12b-it-semancer":{"id":"gemma-4-12b-it-semancer","name":"Gemma 4 12B Semancer","description":"Gemma 4 12B Semancer is an open-weight roleplay finetune with image understanding, tool calling, optional reasoning, and a 131,072-token context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.025}},"Gemma-4-31B-MeroMero-v2":{"id":"Gemma-4-31B-MeroMero-v2","name":"Gemma 4 31B MeroMero v2","description":"Gemma 4 31B MeroMero v2 is a LoRA finetune for emotive dialogue, relationship scenes, creative writing, and multimodal roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemma-4-26b-a4b-it-moonlight":{"id":"gemma-4-26b-a4b-it-moonlight","name":"Moonlight Dusk","description":"Moonlight Dusk is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemma-4-26b-a4b-it-luminous":{"id":"gemma-4-26b-a4b-it-luminous","name":"Luminous Mirror","description":"Luminous Mirror is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"glm-4.1v-thinking-flashx":{"id":"glm-4.1v-thinking-flashx","name":"GLM 4.1V Thinking FlashX","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":64000,"input":64000,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"Gemma-4-31B-GarnetV2":{"id":"Gemma-4-31B-GarnetV2","name":"Gemma 4 31B Garnet V2","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"kimi-k2-instruct-fast":{"id":"kimi-k2-instruct-fast","name":"Kimi K2 0711 Fast","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"status":"deprecated","cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"pokee-isaac":{"id":"pokee-isaac","name":"Pokee-Isaac 28B","description":"Pokee-Isaac is a 28B agentic model with a roughly 10-million-token context window, function calling, and OpenAI-compatible structured output. Pokee bills in $0.01 increments, rounding each non-zero request up to the next cent.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-08-04","last_updated":"2026-08-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"input":10000000,"output":60000},"cost":{"input":0.15,"output":1,"cache_read":0.075}},"gemini-2.5-pro-preview-05-06":{"id":"gemini-2.5-pro-preview-05-06","name":"Gemini 2.5 Pro Preview 0506","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-05-06","last_updated":"2025-05-06","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25}},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":0.375},"canonical_model_id":"google/gemini-2.5-pro"},"gemini-2.5-flash-preview-04-17":{"id":"gemini-2.5-flash-preview-04-17","name":"Gemini 2.5 Flash Preview","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-04-17","last_updated":"2025-04-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.015}},"claw-high":{"id":"claw-high","name":"Claw High","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"doubao-seed-2-0-lite-260215":{"id":"doubao-seed-2-0-lite-260215","name":"Doubao Seed 2.0 Lite","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32000},"cost":{"input":0.1462,"output":0.8738,"cache_read":0.0731},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"celeris-1":{"id":"celeris-1","name":"Celeris 1","description":"Celeris 1 is a diffusion language model built for ultra-low-latency classification, extraction, judging, query rewriting, and other short structured responses.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-07-25","last_updated":"2026-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":2,"output":6,"cache_read":1}},"doubao-seed-2-0-pro-260215":{"id":"doubao-seed-2-0-pro-260215","name":"Doubao Seed 2.0 Pro","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":0.782,"output":3.876,"cache_read":0.391},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"Qwen3.5-27B-Queen-Derestricted":{"id":"Qwen3.5-27B-Queen-Derestricted","name":"Qwen3.5 27B Queen Derestricted","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"Gemma-4-31B-Claude-4.6-Opus-Reasoning-Distilled":{"id":"Gemma-4-31B-Claude-4.6-Opus-Reasoning-Distilled","name":"Gemma 4 31B Claude 4.6 Opus Reasoning Distilled","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"claude","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.0306}},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"Gemini 3 Pro Image","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"input":65536,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"ernie-5.1:thinking":{"id":"ernie-5.1:thinking","name":"ERNIE 5.1 Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"ernie","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2026-05-10","last_updated":"2026-05-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":119000,"input":119000,"output":64000},"cost":{"input":0.75,"output":3,"cache_read":0.75}},"gemma-4-31b-it-scarlet":{"id":"gemma-4-31b-it-scarlet","name":"Gemma 4 31B Scarlet","description":"Scarlet is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-08","last_updated":"2026-10-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemini-2.5-flash-preview-04-17:thinking":{"id":"gemini-2.5-flash-preview-04-17:thinking","name":"Gemini 2.5 Flash Preview Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-04-17","last_updated":"2025-04-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":3.5,"cache_read":0.015}},"gemma-4-26b-a4b-it-darksoul":{"id":"gemma-4-26b-a4b-it-darksoul","name":"Dark Soul","description":"Dark Soul is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemma-4-31b-it-fabled":{"id":"gemma-4-31b-it-fabled","name":"Fabled","description":"Fabled is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"auto-model-premium":{"id":"auto-model-premium","name":"Auto model (Premium)","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":9.996,"output":19.992,"cache_read":4.998}},"gemma-4-31b-it-gembrain":{"id":"gemma-4-31b-it-gembrain","name":"Gembrain","description":"Gembrain is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"holo3-35b-a3b":{"id":"holo3-35b-a3b","name":"Holo3-35B-A3B","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.25,"output":1.8,"cache_read":0.125}},"deepseek-chat-cheaper":{"id":"deepseek-chat-cheaper","name":"DeepSeek V3/Chat Cheaper","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.1,"output":0.425,"cache_read":0.05}},"doubao-seed-2-0-mini-260215":{"id":"doubao-seed-2-0-mini-260215","name":"Doubao Seed 2.0 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32000},"cost":{"input":0.0493,"output":0.4845,"cache_read":0.02465},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"sarvam-105b":{"id":"sarvam-105b","name":"Sarvam 105B","description":"Flagship Indian-language reasoning model for enterprise multilingual applications","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":4096},"cost":{"input":0.054,"output":0.2124,"cache_read":0.0336},"canonical_model_id":"sarvam/sarvam-105b"},"gemini-2.0-pro-reasoner":{"id":"gemini-2.0-pro-reasoner","name":"Gemini 2.0 Pro Reasoner","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-02-05","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":1.292,"output":4.998,"cache_read":0.323}},"longcat-2.5-preview":{"id":"longcat-2.5-preview","name":"LongCat 2.5 Preview","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.75,"output":2.95,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.5-preview"},"Gemma-4-26B-A4B-MeroMero":{"id":"Gemma-4-26B-A4B-MeroMero","name":"Gemma 4 26B A4B MeroMero","description":"Gemma 4 26B A4B MeroMero is an NVFP4 multimodal mixture-of-experts fine-tune for emotive dialogue, relationship scenes, creative writing, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"Qwen3.5-27B-BlueStar-v3-Derestricted":{"id":"Qwen3.5-27B-BlueStar-v3-Derestricted","name":"Qwen3.5 27B BlueStar v3 Derestricted","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"auto-model-standard":{"id":"auto-model-standard","name":"Auto model (Standard)","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":9.996,"output":19.992,"cache_read":4.998}},"command-a-reasoning-08-2025":{"id":"command-a-reasoning-08-2025","name":"Cohere Command A (08/2025)","description":"Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows","family":"command-a","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":8192},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"cohere/command-a-reasoning-08-2025"},"deepseek-v3-0324":{"id":"deepseek-v3-0324","name":"DeepSeek Chat 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.2,"output":0.77,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3-0324"},"longcat-2.0":{"id":"longcat-2.0","name":"LongCat 2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.75,"output":3,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"gemma-4-31b-it-garnet":{"id":"gemma-4-31b-it-garnet","name":"Garnet","description":"Garnet is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"glm-4-long":{"id":"glm-4-long","name":"GLM-4 Long","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":4095},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003}},"mercury-coder-small":{"id":"mercury-coder-small","name":"Mercury Coder Small","description":"Model by Inception AI. A diffusion large language model that runs incredibly quickly (500+ tokens/second) while matching Claude 3.5 Haiku and GPT-4o-mini. 1st in speed on Copilot arena, and matching 2nd in quality.","family":"mercury","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.25,"output":1,"cache_read":0.125}},"longcat-2.0:thinking":{"id":"longcat-2.0:thinking","name":"LongCat 2.0 Thinking","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.75,"output":3,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"gemini-2.5-flash-preview-05-20:thinking":{"id":"gemini-2.5-flash-preview-05-20:thinking","name":"Gemini 2.5 Flash 0520 Thinking","description":"Preview snapshot of Gemini 2.5 Flash with controllable thinking budgets","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"input":1048000,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":3.5,"cache_read":0.015},"canonical_model_id":"google/gemini-2.5-flash-preview-05-20"},"hermes-medium":{"id":"hermes-medium","name":"Hermes Medium","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"auto-model":{"id":"auto-model","name":"Auto model","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":0,"output":0}},"gemini-2.5-pro-preview-03-25":{"id":"gemini-2.5-pro-preview-03-25","name":"Gemini 2.5 Pro Preview 0325","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25}},"phi-4-multimodal-instruct":{"id":"phi-4-multimodal-instruct","name":"Phi 4 Multimodal","description":"Compact Microsoft multimodal model for text, image, and speech understanding","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.07,"output":0.11,"cache_read":0.035},"canonical_model_id":"microsoft/phi-4-multimodal-instruct"},"gemma-4-26b-a4b-it-opusdistill":{"id":"gemma-4-26b-a4b-it-opusdistill","name":"Opus Distill","description":"Opus Distill is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"claw-medium":{"id":"claw-medium","name":"Claw Medium","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"qwen3-vl-235b-a22b-instruct-original":{"id":"qwen3-vl-235b-a22b-instruct-original","name":"Qwen3 VL 235B A22B Instruct Original","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.3,"output":1.9,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"universal-summarizer":{"id":"universal-summarizer","name":"Universal Summarizer","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-23","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":30,"output":30}},"hermes-high":{"id":"hermes-high","name":"Hermes High","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"gemma-4-26b-a4b-it-musica":{"id":"gemma-4-26b-a4b-it-musica","name":"Musica","description":"Musica is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"doubao-1.5-pro-256k":{"id":"doubao-1.5-pro-256k","name":"Doubao 1.5 Pro 256k","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":16384},"cost":{"input":0.799,"output":1.445,"cache_read":0.3995}},"mercury-2":{"id":"mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"qvq-max":{"id":"qvq-max","name":"Qwen: QvQ Max","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-03-28","last_updated":"2025-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":1.2,"output":4.8,"cache_read":0.6}},"gemini-2.5-flash-preview-09-2025":{"id":"gemini-2.5-flash-preview-09-2025","name":"Gemini 2.5 Flash Preview (09/2025)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"ernie-5.1":{"id":"ernie-5.1","name":"ERNIE 5.1","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"ernie","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-05-10","last_updated":"2026-05-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":119000,"input":119000,"output":64000},"cost":{"input":0.75,"output":3,"cache_read":0.75}},"agnes-3.0-flash":{"id":"agnes-3.0-flash","name":"Agnes 3.0 Flash","description":"Agnes 3.0 Flash is a low-cost model for coding, tool use, and multi-turn agent tasks. It supports text and image input, optional thinking, and a 512K-token context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":65536},"cost":{"input":0.05,"output":0.15,"cache_read":0.005}},"gemini-2.5-pro-preview-06-05":{"id":"gemini-2.5-pro-preview-06-05","name":"Gemini 2.5 Pro Preview 0605","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25},"canonical_model_id":"google/gemini-2.5-pro-preview"},"claw-low":{"id":"claw-low","name":"Claw Low","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2,"cache_write":0.08333}},"Gemma-4-31B-DarkIdol":{"id":"Gemma-4-31B-DarkIdol","name":"Gemma 4 31B DarkIdol","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"gemini-2.5-flash-preview-05-20":{"id":"gemini-2.5-flash-preview-05-20","name":"Gemini 2.5 Flash 0520","description":"Preview snapshot of Gemini 2.5 Flash with controllable thinking budgets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"input":1048000,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemini-2.5-flash-preview-05-20"},"gemini-exp-1206":{"id":"gemini-exp-1206","name":"Gemini 2.0 Pro 1206","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":2097152,"input":2097152,"output":8192},"cost":{"input":1.258,"output":4.998,"cache_read":0.629}},"hermes-low":{"id":"hermes-low","name":"Hermes Low","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2,"cache_write":0.08333}},"phi-4-mini-instruct":{"id":"phi-4-mini-instruct","name":"Phi 4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.17,"output":0.68,"cache_read":0.085},"canonical_model_id":"microsoft/phi-4-mini"},"glm-z1-airx":{"id":"glm-z1-airx","name":"GLM Z1 AirX","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":16384},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"Meta-Llama-3-1-8B-Instruct-FP8":{"id":"Meta-Llama-3-1-8B-Instruct-FP8","name":"Llama 3.1 8B (decentralized)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.02,"output":0.03,"cache_read":0.01}},"mistral-code-latest":{"id":"mistral-code-latest","name":"Mistral Code Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.15}},"gemini-2.0-pro-exp-02-05":{"id":"gemini-2.0-pro-exp-02-05","name":"Gemini 2.0 Pro 0205","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-05","last_updated":"2025-02-05","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":2097152,"input":2097152,"output":8192},"cost":{"input":1.989,"output":7.956,"cache_read":0.49725}},"fastgpt":{"id":"fastgpt","name":"Web Answer","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-23","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":7.5,"output":7.5}},"glm-4-plus-0111":{"id":"glm-4-plus-0111","name":"GLM 4 Plus 0111","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":9.996,"output":9.996,"cache_read":4.998}},"ernie-5.0-thinking-preview":{"id":"ernie-5.0-thinking-preview","name":"Ernie 5.0 Thinking Preview","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":1,"output":3.5,"cache_read":0.5}},"Gemma-4-31B-Queen":{"id":"Gemma-4-31B-Queen","name":"Gemma 4 31B Queen","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"holo3-35b-a3b:thinking":{"id":"holo3-35b-a3b:thinking","name":"Holo3-35B-A3B Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.25,"output":1.8,"cache_read":0.125}},"doubao-seed-2-0-code-preview-260215":{"id":"doubao-seed-2-0-code-preview-260215","name":"Doubao Seed 2.0 Code Preview","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":0.782,"output":3.893,"cache_read":0.391},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"gemma-4-31b-it-isometry":{"id":"gemma-4-31b-it-isometry","name":"Isometry","description":"Isometry is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"glm-4.1v-thinking-flash":{"id":"glm-4.1v-thinking-flash","name":"GLM 4.1V Thinking Flash","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":64000,"input":64000,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"MiniMax-M1":{"id":"MiniMax-M1","name":"MiniMax M1","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-01-08","last_updated":"2025-06-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.1394,"output":1.3328,"cache_read":0.0697}},"gemma-4-26b-a4b-it-chimeraxheretic":{"id":"gemma-4-26b-a4b-it-chimeraxheretic","name":"Gemma 4 26B A4B Chimera X Heretic","description":"Chimera X Heretic is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-08","last_updated":"2026-10-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"deepseek-chat":{"id":"deepseek-chat","name":"DeepSeek V3/Deepseek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.1,"output":0.425,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-chat"},"Gemma-4-31B-Cognitive-Unshackled":{"id":"Gemma-4-31B-Cognitive-Unshackled","name":"Gemma 4 31B Cognitive Unshackled","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"gemini-2.5-flash-preview-09-2025-thinking":{"id":"gemini-2.5-flash-preview-09-2025-thinking","name":"Gemini 2.5 Flash Preview (09/2025) – Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"gemma-4-31b-it-gemsicle":{"id":"gemma-4-31b-it-gemsicle","name":"Gemsicle","description":"Gemsicle is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemma-4-26b-a4b-it-shadowsiren":{"id":"gemma-4-26b-a4b-it-shadowsiren","name":"Shadow Siren","description":"Shadow Siren is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemini-2.5-pro-exp-03-25":{"id":"gemini-2.5-pro-exp-03-25","name":"Gemini 2.5 Pro Experimental 0325","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25}},"asi1-mini":{"id":"asi1-mini","name":"ASI1 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":1,"output":1,"cache_read":0.5}},"Gemma-4-31B-MeroMero-v2:thinking":{"id":"Gemma-4-31B-MeroMero-v2:thinking","name":"Gemma 4 31B MeroMero v2 Thinking","description":"Gemma 4 31B MeroMero v2 with thinking enabled for more deliberate emotive dialogue, relationship scenes, creative writing, and multimodal roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemma-4-12b-it":{"id":"gemma-4-12b-it","name":"Gemma 4 12B Instruct","description":"Compact Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.025},"canonical_model_id":"google/gemma-4-12b-it"},"GLM-4.6-Derestricted-v5":{"id":"GLM-4.6-Derestricted-v5","name":"GLM 4.6 Derestricted v5","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":8192},"cost":{"input":0.4,"output":1.5,"cache_read":0.2}},"gemma-4-31b-it-darkidol":{"id":"gemma-4-31b-it-darkidol","name":"DarkIdol","description":"DarkIdol is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemini-2.5-flash-lite-preview-09-2025-thinking":{"id":"gemini-2.5-flash-lite-preview-09-2025-thinking","name":"Gemini 2.5 Flash Lite Preview (09/2025) – Thinking","description":"September 2025 preview of Gemini 2.5 Flash-Lite with better instruction following and multimodal quality","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-09-2025"},"auto-model-basic":{"id":"auto-model-basic","name":"Auto model (Basic)","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":9.996,"output":19.992,"cache_read":4.998}},"gemma-4-26b-a4b-it-chimerax":{"id":"gemma-4-26b-a4b-it-chimerax","name":"Chimera X","description":"Chimera X is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemma-4-31b-it-novelist":{"id":"gemma-4-31b-it-novelist","name":"Novelist","description":"Novelist is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite"},"meganova-ai/manta-mini-1.0":{"id":"meganova-ai/manta-mini-1.0","name":"Manta Mini 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-20","last_updated":"2025-12-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":0.02,"output":0.16,"cache_read":0.01}},"meganova-ai/manta-flash-1.0":{"id":"meganova-ai/manta-flash-1.0","name":"Manta Flash 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-20","last_updated":"2025-12-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"input":16384,"output":16384},"cost":{"input":0.02,"output":0.16,"cache_read":0.01}},"meganova-ai/manta-pro-1.0":{"id":"meganova-ai/manta-pro-1.0","name":"Manta Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-20","last_updated":"2025-12-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"input":65536,"output":32768},"cost":{"input":0.06,"output":0.5,"cache_read":0.03}},"MarinaraSpaghetti/NemoMix-Unleashed-12B":{"id":"MarinaraSpaghetti/NemoMix-Unleashed-12B","name":"NemoMix 12B Unleashed","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":80000},"cost":{"input":0.25,"output":0.9,"cache_read":0.125},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"lightonai/LightOnOCR-2-1B":{"id":"lightonai/LightOnOCR-2-1B","name":"LightOnOCR 2","description":"LightOnOCR 2 hosted by IONOS in Berlin, Germany. Zero data retention.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.1785,"output":0.3465}},"abacusai/Dracarys-72B-Instruct":{"id":"abacusai/Dracarys-72B-Instruct","name":"Llama 3.1 70B Dracarys 2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-08-02","last_updated":"2025-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"deepseek/deepseek-v4.1-flash-1017:thinking":{"id":"deepseek/deepseek-v4.1-flash-1017:thinking","name":"DeepSeek V4.1 Flash 1017 Thinking","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":393216},"cost":{"input":0.1,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":0.0537,"output":0.2148,"cache_read":0.001074},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-latest":{"id":"deepseek/deepseek-latest","name":"DeepSeek Latest","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.132,"output":0.396,"cache_read":0.0042}},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.7,"output":1.4,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash-0731:thinking":{"id":"deepseek/deepseek-v4-flash-0731:thinking","name":"DeepSeek V4 Flash 0731 (Thinking)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.04444,"output":0.13332,"cache_read":0.001414},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash-0423:thinking":{"id":"deepseek/deepseek-v4-flash-0423:thinking","name":"DeepSeek V4 Flash 0423 (Thinking)","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-v4-flash:thinking":{"id":"deepseek/deepseek-v4-flash:thinking","name":"DeepSeek V4 Flash (Thinking)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v3.2:thinking":{"id":"deepseek/deepseek-v3.2:thinking","name":"DeepSeek V3.2 Thinking","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":163000,"input":163000,"output":65536},"cost":{"input":0.2088,"output":0.3096,"cache_read":0.0216},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":393216},"cost":{"input":0.1,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.132,"output":0.396,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-pro-0813:thinking":{"id":"deepseek/deepseek-v4-pro-0813:thinking","name":"DeepSeek V4 Pro 0813 Thinking","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.132,"output":0.396,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.00686},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-v4.1-flash:thinking":{"id":"deepseek/deepseek-v4.1-flash:thinking","name":"DeepSeek V4.1 Flash Thinking","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":0.0537,"output":0.2148,"cache_read":0.001074},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.04444,"output":0.13332,"cache_read":0.001414},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash-0423":{"id":"deepseek/deepseek-v4-flash-0423","name":"DeepSeek V4 Flash 0423","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-v4-flash-latest":{"id":"deepseek/deepseek-v4-flash-latest","name":"DeepSeek V4 Flash Latest","description":"Compatibility alias that routes to the newest dated DeepSeek V4 Flash release. Currently routes to DeepSeek V4 Flash 0731. ⚠️ This route goes directly to DeepSeek, so privacy and logging guarantees are limited.","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-08-02","last_updated":"2026-08-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":393216},"cost":{"input":0.1,"output":0.6,"cache_read":0.003}},"deepseek/deepseek-v4-pro:thinking":{"id":"deepseek/deepseek-v4-pro:thinking","name":"DeepSeek V4 Pro (Thinking)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.7,"output":1.4,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":163000,"input":163000,"output":65536},"cost":{"input":0.2088,"output":0.3096,"cache_read":0.0216},"canonical_model_id":"deepseek/deepseek-v3.2"},"baseten/Kimi-K2-Instruct-FP4":{"id":"baseten/Kimi-K2-Instruct-FP4","name":"Kimi K2 0711 Instruct FP4","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":1.8,"cache_read":0.2}},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":16384},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"sakana/fugu-ultra-v1.1":{"id":"sakana/fugu-ultra-v1.1","name":"Fugu Ultra v1.1","description":"Sakana AI's upgraded Fugu Ultra release with stronger coding, agentic task execution, and advanced reasoning through dynamic orchestration of frontier models.","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":16384},"cost":{"input":5,"output":30,"cache_read":0.5}},"inception/mercury-2.5-preview":{"id":"inception/mercury-2.5-preview","name":"Mercury 2.5 Preview","description":"Mercury 2.5 Preview is Inception's latest and most intelligent diffusion language model. Instead of generating tokens strictly one at a time, it produces and refines multiple tokens in parallel, reaching up to 1,107 tokens per second on standard GPUs. It delivers a 10+ point intelligence gain over Mercury 2, with tunable reasoning, parallel tool calls, schema-aligned JSON output, and a 260K context window. It is built for latency-sensitive production work such as search agents, voice pipelines, customer support, rapid coding iteration, and coding subagents.","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"input":260000,"output":65536},"cost":{"input":0.04,"output":0.15,"cache_read":0.004}},"inception/mercury-3":{"id":"inception/mercury-3","name":"Mercury 3","description":"Inception's diffusion reasoning model for chat, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01},"canonical_model_id":"inception/mercury-3"},"ateron/gemma-4-dark-thoughts-v2-31b":{"id":"ateron/gemma-4-dark-thoughts-v2-31b","name":"Gemma 4 Dark Thoughts V2 31B","description":"Ateron's second Dark Thoughts merge combines Dark Scarlett V2, MeroMero V2, and Scotoma 2 for creative dialogue and extended character roleplay. This text-only Gemma 4 31B deployment supports a 65,536-token context window and up to 8,192 output tokens.","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"Sao10K/L3.1-70B-Euryale-v2.2":{"id":"Sao10K/L3.1-70B-Euryale-v2.2","name":"Llama 3.1 70B Euryale","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":20480,"input":20480,"output":16384},"cost":{"input":0.306,"output":0.357,"cache_read":0.153},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"Sao10K/L3.3-70B-Euryale-v2.3":{"id":"Sao10K/L3.3-70B-Euryale-v2.3","name":"Llama 3.3 70B Euryale","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-12-18","last_updated":"2024-12-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":20480,"input":20480,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465},"canonical_model_id":"sao10k/l3.3-euryale-70b"},"Sao10K/L3-8B-Stheno-v3.2":{"id":"Sao10K/L3-8B-Stheno-v3.2","name":"Sao10K Stheno 8b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-11-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003}},"Sao10K/L3.1-70B-Hanami-x1":{"id":"Sao10K/L3.1-70B-Hanami-x1","name":"Llama 3.1 70B Hanami","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"VongolaChouko/Starcannon-Unleashed-12B-v1.0":{"id":"VongolaChouko/Starcannon-Unleashed-12B-v1.0","name":"Mistral Nemo Starcannon 12b v1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"aion-labs/aion-3.5":{"id":"aion-labs/aion-3.5","name":"AionLabs: Aion 3.5","description":"A GLM-family collaborative generation model tuned for immersive roleplay and storytelling, with stronger narrative structure, tension, conflict, and nuanced mature themes.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75}},"aion-labs/aion-2.0":{"id":"aion-labs/aion-2.0","name":"AionLabs: Aion-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-21","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.8,"output":1.6,"cache_read":0.2},"canonical_model_id":"aion-labs/aion-2.0"},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"Llama 3.1 8b (uncensored)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.8,"output":1.6,"cache_read":0.4},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"aion-labs/aion-3.0-mini":{"id":"aion-labs/aion-3.0-mini","name":"AionLabs: Aion 3.0 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18},"canonical_model_id":"aion-labs/aion-3.0-mini"},"aion-labs/aion-3.0":{"id":"aion-labs/aion-3.0","name":"AionLabs: Aion 3.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75},"canonical_model_id":"aion-labs/aion-3.0"},"aion-labs/aion-3.5-mini":{"id":"aion-labs/aion-3.5-mini","name":"AionLabs: Aion 3.5 Mini","description":"A GLM-family collaborative generation model tuned for immersive roleplay and storytelling, with stronger narrative structure, tension, conflict, and nuanced mature themes.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18}},"MiniMaxAI/MiniMax-M1-80k":{"id":"MiniMaxAI/MiniMax-M1-80k","name":"MiniMax M1 80K","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.6052,"output":2.4225,"cache_read":0.3026},"canonical_model_id":"minimax/minimax-m1-80k"},"moondream/moondream3-preview":{"id":"moondream/moondream3-preview","name":"Moondream 3 Preview","description":"Moondream 3 Preview is the original 9B mixture-of-experts vision-language checkpoint with 2B active parameters, for visual question answering, captioning, OCR, and visual reasoning. It accepts text or one image with optional reasoning, a 32,000-token context, and up to 4,096 output tokens.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"moondream/moondream3.1-9b-a2b":{"id":"moondream/moondream3.1-9b-a2b","name":"Moondream 3.1 9B A2B","description":"Moondream 3.1 is the successor to Moondream 3 Preview: a distinct 9B mixture-of-experts vision-language checkpoint with 2B active parameters, for visual question answering, captioning, OCR, and visual reasoning. It accepts text or one image with optional reasoning, a 32,000-token context, and up to 4,096 output tokens.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"nanogpt/coding-router:low":{"id":"nanogpt/coding-router:low","name":"Coding Router Low","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028}},"nanogpt/coding-router:max":{"id":"nanogpt/coding-router:max","name":"Coding Router Max","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5}},"nanogpt/coding-router":{"id":"nanogpt/coding-router","name":"Coding Router","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":1.1,"output":2.2,"cache_read":0.11}},"nanogpt/coding-router:high":{"id":"nanogpt/coding-router:high","name":"Coding Router High","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":1.1,"output":2.2,"cache_read":0.11}},"nanogpt/coding-router:medium":{"id":"nanogpt/coding-router:medium","name":"Coding Router Medium","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028}},"reflection/beam-1.0":{"id":"reflection/beam-1.0","name":"Beam 1.0 501B A23B","description":"Reflection's Apache 2.0 open-weight text model with 501.3B total and 23.4B active parameters, built for coding, reasoning and tool use. It supports a shared 1,048,576-token context window, up to 131,072 output tokens, JSON mode and structured output. Reasoning is always enabled with low, medium, high, xhigh and max effort; NanoGPT defaults to max. Knowledge cutoff: June 30, 2026.","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"LatitudeGames/Wayfarer-Large-70B-Llama-3.3":{"id":"LatitudeGames/Wayfarer-Large-70B-Llama-3.3","name":"Llama 3.3 70B Wayfarer","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-20","last_updated":"2025-02-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"Gryphe/MythoMax-L2-13b":{"id":"Gryphe/MythoMax-L2-13b","name":"MythoMax 13B","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"input":4096,"output":3686},"cost":{"input":0.1003,"output":0.1003,"cache_read":0.05015},"canonical_model_id":"gryphe/mythomax-l2-13b"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":900000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.20-multi-agent":{"id":"x-ai/grok-4.20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"input":2000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":230400},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"input":2000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-latest":{"id":"x-ai/grok-latest","name":"Grok Latest","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5}},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"nimbz/gemma-4-giftige-blume-31b-v2":{"id":"nimbz/gemma-4-giftige-blume-31b-v2","name":"Gemma 4 Giftige-Blume V2 31B","description":"Nimbz's second Giftige-Blume merge focuses on character roleplay, creative prose, and stricter prompt adherence. It combines storymaxxed, Equinox, Mero-Artemis, Musica, Ortenzya, Gemopus, and Sprinkle. This text-only Gemma 4 31B deployment supports a 65,536-token context window and up to 8,192 output tokens.","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"nimbz/schattenblume-31b":{"id":"nimbz/schattenblume-31b","name":"Schattenblume 31B","description":"Nimbz's Gemma 4 31B merge for character roleplay, expressive prose, and creative dialogue. It combines Scotoma 2, Giftige-Blume v1, and MeroMero v2. This text-only deployment supports a 65,536-token context window and up to 8,192 output tokens.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"heabsy/qwen3.8-27b-uncensored-fast":{"id":"heabsy/qwen3.8-27b-uncensored-fast","name":"Qwen 3.8 27B Uncensored Fast by Heabsy","description":"Heabsy's NVFP4 build of Qwen3.8-27B by Alibaba Cloud, with refusal removal based on huihui-ai's Huihui-Qwen3.8-27B-abliterated. Quantized and calibrated by Heabsy, with unpublished weights under the Heabsy Model License. Supports text and image input, tools, and optional reasoning capped at 16,384 tokens within the output budget. This named build is separate from Qwen 3.8 27B Uncensored.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":235929},"cost":{"input":1,"output":3,"cache_read":0.22}},"heabsy/cyberheabsy":{"id":"heabsy/cyberheabsy","name":"CyberHeabsy","description":"A low-refusal text model for authorized security research, roleplay, and creative writing. It preserves tool calling after refusal removal, supports optional reasoning, and has a 1M-token context window.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-27","last_updated":"2026-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.5,"output":4.5,"cache_read":0.275}},"Doctor-Shotgun/MS3.2-24B-Magnum-Diamond":{"id":"Doctor-Shotgun/MS3.2-24B-Magnum-Diamond","name":"MS3.2 24B Magnum Diamond","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"thinkingmachines/Inkling-Small:thinking":{"id":"thinkingmachines/Inkling-Small:thinking","name":"Inkling Small Thinking","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling:thinking":{"id":"thinkingmachines/inkling:thinking","name":"Inkling Thinking","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"input":1048000,"output":32768},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"input":1048000,"output":32768},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"reka/reka-edge-2603":{"id":"reka/reka-edge-2603","name":"Reka Edge 2603","description":"Reka's compact multimodal model for visual understanding and tool use, served directly by Reka.","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":16384},"cost":{"input":0.105,"output":0.105},"canonical_model_id":"rekaai/reka-edge"},"reka/reka-flash-3":{"id":"reka/reka-flash-3","name":"Reka Flash 3","description":"Reka's 21B reasoning model for chat and coding, served directly by Reka.","family":"reka","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":65536},"cost":{"input":0.105,"output":0.21},"canonical_model_id":"rekaai/reka-flash-3"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.1-flash"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.0-flash:thinking":{"id":"inclusionai/ling-3.0-flash:thinking","name":"Ling 3.0 Flash Thinking","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash"},"THUDM/GLM-Z1-9B-0414":{"id":"THUDM/GLM-Z1-9B-0414","name":"GLM Z1 9B 0414","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm-z","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8000},"cost":{"input":0.2,"output":0.2,"cache_read":0.1}},"THUDM/GLM-4-9B-0414":{"id":"THUDM/GLM-4-9B-0414","name":"GLM 4 9B 0414","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8000},"cost":{"input":0.2,"output":0.2,"cache_read":0.1}},"THUDM/GLM-4-32B-0414":{"id":"THUDM/GLM-4-32B-0414","name":"GLM 4 32B 0414","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.2,"output":0.2,"cache_read":0.1}},"shisa-ai/shisa-v2.1-llama3.3-70b":{"id":"shisa-ai/shisa-v2.1-llama3.3-70b","name":"Shisa V2.1 Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":4096},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"shisa-ai/shisa-v2-llama3.3-70b":{"id":"shisa-ai/shisa-v2-llama3.3-70b","name":"Shisa V2 Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"minimax/minimax-m3:thinking":{"id":"minimax/minimax-m3:thinking","name":"MiniMax M3 Thinking","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"input":512000,"output":80000},"cost":{"input":0.23,"output":0.96,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":0.27,"output":1.08,"cache_read":0.027},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":0.21,"output":0.84,"cache_read":0.042},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.33,"output":1.32,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-01":{"id":"minimax/minimax-01","name":"MiniMax 01","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-01-15","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000192,"input":1000192,"output":16384},"cost":{"input":0.1394,"output":1.122,"cache_read":0.0697}},"minimax/minimax-latest":{"id":"minimax/minimax-latest","name":"MiniMax Latest","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"input":512000,"output":80000},"cost":{"input":0.23,"output":0.96,"cache_read":0.05}},"minimax/minimax-m2.7-turbo":{"id":"minimax/minimax-m2.7-turbo","name":"MiniMax M2.7 Turbo","description":"Efficient MiniMax model for quick assistance, coding, and routine automation","family":"minimax-m2.7","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"input":512000,"output":80000},"cost":{"input":0.23,"output":0.96,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.17,"output":1.53,"cache_read":0.085},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m2-her":{"id":"minimax/minimax-m2-her","name":"MiniMax M2-her","description":"MiniMax M2 variant tuned for conversational and character-driven agent interactions","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65532,"input":65532,"output":2048},"cost":{"input":0.302,"output":1.207,"cache_read":0.151},"canonical_model_id":"minimax/MiniMax-M2-Her"},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B","description":"IBM Granite 4.2 8B is an Apache 2.0-licensed dense model with native step-by-step reasoning and specialized training for agentic work. It can plan before acting, sequence tools, navigate codebases, work in terminals, and verify results across coding, search, mathematics, science, and complex instruction-following tasks.","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.1,"output":0.15,"cache_read":0.05},"canonical_model_id":"ibm/granite-4.2-8b"},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"input":6144,"output":4096},"cost":{"input":0.799,"output":1.207,"cache_read":0.3995},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Cohere: Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":2.856,"output":14.246,"cache_read":1.428},"canonical_model_id":"cohere/command-r-plus-08-2024"},"Salesforce/Llama-xLAM-2-70b-fc-r":{"id":"Salesforce/Llama-xLAM-2-70b-fc-r","name":"Llama-xLAM-2 70B fc-r","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-04-13","last_updated":"2025-04-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":2.5,"cache_read":1.25}},"bytedance-seed/seed-2-1-turbo":{"id":"bytedance-seed/seed-2-1-turbo","name":"ByteDance Seed 2.1 Turbo","description":"ByteDance Seed 2.1 Turbo is a multimodal model for coding and long-horizon agent workflows, including end-to-end software delivery and multi-step task execution. It supports text, image, and video input with a 262k context window.","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":235929},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"ByteDance Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.25,"output":2,"cache_read":0.125},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"ByteDance Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.25},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"microsoft/wizardlm-2-8x22b":{"id":"microsoft/wizardlm-2-8x22b","name":"WizardLM-2 8x22B","description":"Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-04-15","last_updated":"2024-04-15","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465},"canonical_model_id":"microsoft/wizardlm-2-8x22b"},"Envoid/Llama-3.05-NT-Storybreaker-Ministral-70B":{"id":"Envoid/Llama-3.05-NT-Storybreaker-Ministral-70B","name":"Llama 3.05 Storybreaker Ministral 70b","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Envoid/Llama-3.05-Nemotron-Tenyxchat-Storybreaker-70B":{"id":"Envoid/Llama-3.05-Nemotron-Tenyxchat-Storybreaker-70B","name":"Nemotron Tenyxchat Storybreaker 70b","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.01},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-s-2.1:thinking":{"id":"poolside/laguna-s-2.1:thinking","name":"Laguna S 2.1 Thinking","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.01},"canonical_model_id":"poolside/laguna-s-2.1"},"failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5":{"id":"failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5","name":"Llama 3 70B abliterated","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"xiaomi/mimo-v2.5-pro:thinking":{"id":"xiaomi/mimo-v2.5-pro:thinking","name":"MiMo V2.5 Pro Thinking","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-flash-uncensored":{"id":"xiaomi/mimo-v2.6-flash-uncensored","name":"MiMo V2.6 Flash Uncensored","description":"A text-only, lower-refusal MiMo V2.6 Flash finetune served by Heabsy with a 1M-token context window, separate reasoning output, and tool calling.","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":1.5,"cache_read":0.1995}},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.6-flash-abliterated":{"id":"xiaomi/mimo-v2.6-flash-abliterated","name":"MiMo V2.6 Flash Abliterated","description":"A MiMo V2.6 Flash finetune with refusal-direction ablation, image input, a 1M-token context window, separate reasoning output, and tool calling.","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":1.5,"cache_read":0.275}},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"xiaomi/mimo-v2.6-flash-uncensored:thinking":{"id":"xiaomi/mimo-v2.6-flash-uncensored:thinking","name":"MiMo V2.6 Flash Uncensored Thinking","description":"MiMo V2.6 Flash Uncensored with maximum thinking enabled. Supports a 1M-token context window and tool calling.","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-26","last_updated":"2026-09-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":1.5,"cache_read":0.1995}},"xiaomi/mimo-v2.5:thinking":{"id":"xiaomi/mimo-v2.5:thinking","name":"MiMo V2.5 Thinking","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"LLM360/K2-Think":{"id":"LLM360/K2-Think","name":"K2-Think","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":0.17,"output":0.68,"cache_read":0.085}},"perceptron/perceptron-mk1":{"id":"perceptron/perceptron-mk1","name":"Perceptron Mk1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.15,"output":1.5,"cache_read":0.075},"canonical_model_id":"perceptron/perceptron-mk1"},"perceptron/perceptron-mk1.5":{"id":"perceptron/perceptron-mk1.5","name":"Perceptron Mk1.5","description":"Perceptron Mk1.5 is a multimodal reasoning model for visual analysis and physical agents. It accepts text, images, video, and audio, supports tool calling and structured output, and can return spatial annotations such as points, boxes, polygons, and tracks.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":36864,"input":36864,"output":8192},"cost":{"input":0.15,"output":1.5}},"NeverSleep/Lumimaid-v0.2-70B":{"id":"NeverSleep/Lumimaid-v0.2-70B","name":"Lumimaid v0.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":1,"output":1.5,"cache_read":0.5}},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-mini4:thinking":{"id":"upstage/solar-mini4:thinking","name":"Solar Mini 4 Thinking","description":"Upstage's compact, cost-efficient model for agentic use","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.005},"canonical_model_id":"upstage/solar-mini4"},"upstage/solar-pro4:thinking":{"id":"upstage/solar-pro4:thinking","name":"Solar Pro 4 Thinking","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-mini4":{"id":"upstage/solar-mini4","name":"Solar Mini 4","description":"Upstage's compact, cost-efficient model for agentic use","family":"solar-mini","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.005},"canonical_model_id":"upstage/solar-mini4"},"aisingapore/Gemma-SEA-LION-v4-27B-IT":{"id":"aisingapore/Gemma-SEA-LION-v4-27B-IT","name":"Gemma SEA-LION v4 27B","description":"Gemma 3 27B tuned by AI Singapore for Southeast Asian languages and instruction following","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":128000},"cost":{"input":0.24,"output":0.48},"canonical_model_id":"aisingapore/gemma-sea-lion-v4-27b-it"},"google/gemma-4-26b-a4b-it-cybersecurity":{"id":"google/gemma-4-26b-a4b-it-cybersecurity","name":"Gemma 4 26B A4B Cybersecurity","description":"Gemma 4 26B A4B Cybersecurity is a cybersecurity-focused variant based on the uncensored model, with provider moderation for illegal activities. It supports optional reasoning, image understanding, tool calling, and a 262,144-token context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-19","last_updated":"2026-09-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1056,"output":0.3344,"cache_read":0.0528}},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-flash-latest"},"google/gemini-3.1-pro-preview-high":{"id":"google/gemini-3.1-pro-preview-high","name":"Gemini 3.1 Pro (Preview High)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-pro-latest":{"id":"google/gemini-pro-latest","name":"Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375}},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro (Preview Custom Tools)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-flash-preview-thinking":{"id":"google/gemini-3-flash-preview-thinking","name":"Gemini 3 Flash Thinking","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.0675,"output":0.225,"cache_read":0.0375},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemma-4-26b-a4b-it:thinking":{"id":"google/gemma-4-26b-a4b-it:thinking","name":"Gemma 4 26B A4B Thinking","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.0675,"output":0.225,"cache_read":0.0375},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-4-argon"},"google/diffusiongemma":{"id":"google/diffusiongemma","name":"DiffusionGemma","description":"DiffusionGemma is a high-speed diffusion-based version of Gemma 4 26B A4B. It supports optional reasoning and a 262,144-token context window.","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","xhigh"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.05,"output":0.15,"cache_read":0.025},"canonical_model_id":"google/diffusiongemma-26b-a4b-it"},"google/gemini-3.1-pro-preview-low":{"id":"google/gemini-3.1-pro-preview-low","name":"Gemini 3.1 Pro (Preview Low)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.5-flash-thinking":{"id":"google/gemini-3.5-flash-thinking","name":"Gemini 3.5 Flash Thinking","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.083333},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemma4-31b-splituntied":{"id":"google/gemma4-31b-splituntied","name":"Gemma 4 31B Split-Untied","description":"Blazed-Forge's Split-Untied is a text-only Gemma 4 31B community finetune with an untied BF16 output head, built for creative writing, roleplay, expressive dialogue, and tool use.","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.3,"cache_read":0.05}},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash (Preview)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.14,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemma-4-31b-it:thinking":{"id":"google/gemma-4-31b-it:thinking","name":"Gemma 4 31B Thinking","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.1,"output":0.34,"cache_read":0.1},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.083333},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-flash-lite-latest"},"GalrionSoftworks/MN-LooseCannon-12B-v1":{"id":"GalrionSoftworks/MN-LooseCannon-12B-v1","name":"MN-LooseCannon-12B-v1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"huihui-ai/DeepSeek-R1-Distill-Llama-70B-abliterated":{"id":"huihui-ai/DeepSeek-R1-Distill-Llama-70B-abliterated","name":"DeepSeek R1 Llama 70B Abliterated","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-01-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"huihui-ai/Llama-3.3-70B-Instruct-abliterated":{"id":"huihui-ai/Llama-3.3-70B-Instruct-abliterated","name":"Llama 3.3 70B Instruct abliterated","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-08-08","last_updated":"2025-08-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"huihui-ai/Qwen2.5-32B-Instruct-abliterated":{"id":"huihui-ai/Qwen2.5-32B-Instruct-abliterated","name":"Qwen 2.5 32B Abliterated","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-01-06","last_updated":"2025-01-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated":{"id":"huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated","name":"DeepSeek R1 Qwen Abliterated","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-01-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":1.4,"output":1.4,"cache_read":0.7}},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"stepfun/step-3.7-flash:thinking":{"id":"stepfun/step-3.7-flash:thinking","name":"Step 3.7 Flash Thinking","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"featherless-ai/Qwerky-72B":{"id":"featherless-ai/Qwerky-72B","name":"Qwerky 72B","description":"General-purpose chat model for instruction following, writing, and analysis","family":"qwerky","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-03-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"abliteration-ai/abliterated-model-large":{"id":"abliteration-ai/abliterated-model-large","name":"Abliterated Model Large","description":"Abliteration.ai's large text reasoning model is derived from GLM-5.2 and supports native tool calling, structured output, automatic prompt caching, and a one-million-token context window.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":3,"output":5,"cache_read":0.3}},"abliteration-ai/abliterated-model":{"id":"abliteration-ai/abliterated-model","name":"Abliterated Model","description":"Abliteration.ai's multimodal reasoning model supports text and image input, structured output, automatic prompt caching, and a 262K-token context window.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":262134},"cost":{"input":1,"output":3,"cache_read":0.1}},"abliteration-ai/abliterated-model-large-v2":{"id":"abliteration-ai/abliterated-model-large-v2","name":"Abliterated Model Large V2","description":"Abliteration.ai's default large text reasoning model is derived from GLM-5.3 for harder reasoning and evaluation workloads, with automatic prompt caching and a one-million-token context window.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-31","last_updated":"2026-08-31","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":3,"output":5,"cache_read":0.3}},"abliteration-ai/abliterated-model-large-flash":{"id":"abliteration-ai/abliterated-model-large-flash","name":"Abliterated Large Flash","description":"Abliterated Large Flash is an uncensored reasoning model based on DeepSeek V4.1-Flash-1017. It accepts text and images and returns text, with tool calling, JSON mode, JSON Schema output, and automatic prompt caching. Its shared input/output context is 1,048,576 tokens, with up to 262,144 output tokens when space permits. Reasoning defaults to high and can be turned off or set to low, high, or max. Supports up to four images across the request history.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":262144},"cost":{"input":1,"output":2,"cache_read":0.1}},"openai/o1-pro":{"id":"openai/o1-pro","name":"OpenAI o1 Pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":150,"output":600,"cache_read":75},"canonical_model_id":"openai/o1-pro"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"input":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT 5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT 5.6 Terra Pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT 5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT 5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT 6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6.1-sol-pro":{"id":"openai/gpt-6.1-sol-pro","name":"GPT 6.1 Sol Pro","description":"GPT-6.1 Sol Pro uses the same underlying model as GPT-6.1 Sol with Pro reasoning mode enabled for higher-quality responses on complex tasks.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/o3-mini":{"id":"openai/o3-mini","name":"OpenAI o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.02,"output":0.1},"canonical_model_id":"openai/gpt-oss-20b"},"openai/o3-mini-low":{"id":"openai/o3-mini-low","name":"OpenAI o3-mini (Low)","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.075,"output":0.3},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT 6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT 4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT 5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT 5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT 5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-chat-latest":{"id":"openai/gpt-chat-latest","name":"GPT Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT 5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":2.5,"output":20,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/o4-mini-high":{"id":"openai/o4-mini-high","name":"OpenAI o4-mini high","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o4-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT 5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-latest":{"id":"openai/gpt-latest","name":"GPT Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT 4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT 5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra-pro":{"id":"openai/gpt-6-astra-pro","name":"GPT 6 Astra Pro","description":"GPT 6 Astra in Pro reasoning mode. Uses additional model work for difficult tasks, with higher latency and token usage at the same per-token rates. Reasoning effort remains independently configurable.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT 6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT 5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":15,"output":120,"cache_read":1.5},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-astra-latest":{"id":"openai/gpt-astra-latest","name":"GPT Astra Latest","description":"Compatibility alias that routes to GPT 6 Astra, the latest supported GPT Astra model.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/gpt-5.1-2025-11-13":{"id":"openai/gpt-5.1-2025-11-13","name":"GPT-5.1 (2025-11-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/o1":{"id":"openai/o1","name":"OpenAI o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT 5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/o3-pro-2025-06-10":{"id":"openai/o3-pro-2025-06-10","name":"OpenAI o3-pro (2025-06-10)","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":22,"output":88,"cache_read":11},"canonical_model_id":"openai/o3-pro"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT 5.6 Sol Pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT 5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-6-sol-pro":{"id":"openai/gpt-6-sol-pro","name":"GPT 6 Sol Pro","description":"GPT-6 Sol Pro uses the same underlying model as GPT-6 Sol with Pro reasoning mode enabled for higher-quality responses on complex tasks.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-6-luna-pro":{"id":"openai/gpt-6-luna-pro","name":"GPT 6 Luna Pro","description":"GPT-6 Luna Pro uses the same underlying model as GPT-6 Luna with Pro reasoning mode enabled for higher-quality responses on complex tasks.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"openai/o4-mini":{"id":"openai/o4-mini","name":"OpenAI o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o4-mini"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT 5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT 5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT 5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-luna-latest":{"id":"openai/gpt-luna-latest","name":"GPT Luna Latest","description":"Compatibility alias that routes to GPT 5.6 Luna, the latest supported GPT Luna model.","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"openai/o3":{"id":"openai/o3","name":"OpenAI o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":1},"canonical_model_id":"openai/o3"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT 5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT 5.6 Luna Pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT 5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-sol-latest":{"id":"openai/gpt-sol-latest","name":"GPT Sol Latest","description":"Compatibility alias that routes to GPT 5.6 Sol, the latest supported GPT Sol model.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"openai/gpt-terra-latest":{"id":"openai/gpt-terra-latest","name":"GPT Terra Latest","description":"Compatibility alias that routes to GPT 5.6 Terra, the latest supported GPT Terra model.","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT 6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT 4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"OpenAI o3-mini (High)","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT 5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"deepseek-ai/DeepSeek-V3.1:thinking":{"id":"deepseek-ai/DeepSeek-V3.1:thinking","name":"DeepSeek V3.1 Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.2,"output":0.7,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.2,"output":0.7,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"input":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V3.1-Terminus:thinking":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus:thinking","name":"DeepSeek V3.1 Terminus (Thinking)","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek-thinking","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2024-01-01","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.25,"output":0.7,"cache_read":0.125}},"deepseek-ai/deepseek-v3.2-exp-thinking":{"id":"deepseek-ai/deepseek-v3.2-exp-thinking","name":"DeepSeek V3.2 Exp Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"input":163840,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.14}},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2025-08-02","last_updated":"2025-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.25,"output":0.7,"cache_read":0.125}},"deepseek-ai/deepseek-v3.2-exp":{"id":"deepseek-ai/deepseek-v3.2-exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"input":163840,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.14}},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Amazon Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"input":300000,"output":5120},"cost":{"input":0.0595,"output":0.238,"cache_read":0.02975},"canonical_model_id":"amazon/nova-lite-v1"},"amazon/nova-2-lite-v1":{"id":"amazon/nova-2-lite-v1","name":"Amazon Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65535},"cost":{"input":0.51,"output":4.25,"cache_read":0.255},"canonical_model_id":"amazon/nova-2-lite-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Amazon Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"input":300000,"output":32000},"cost":{"input":0.799,"output":3.196,"cache_read":0.3995},"canonical_model_id":"amazon/nova-pro-v1"},"ReadyArt/MS3.2-The-Omega-Directive-24B-Unslop-v2.0":{"id":"ReadyArt/MS3.2-The-Omega-Directive-24B-Unslop-v2.0","name":"Omega Directive 24B Unslop v2.0","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"deepcogito/cogito-v1-preview-qwen-32B":{"id":"deepcogito/cogito-v1-preview-qwen-32B","name":"Cogito v1 Preview Qwen 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-05-10","last_updated":"2025-05-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":1.8,"output":1.8,"cache_read":0.9}},"bytedance/doubao-seed-2.1-turbo":{"id":"bytedance/doubao-seed-2.1-turbo","name":"Doubao Seed 2.1 Turbo","description":"Fast, lower-cost model in the Doubao Seed 2.1 family for everyday chat, coding assistance, document work, and high-throughput productivity tasks. Supports a 256k context window and up to 128k output tokens. Note: privacy and logging guarantees may be limited.","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"bytedance/doubao-seed-2.1-pro":{"id":"bytedance/doubao-seed-2.1-pro","name":"Doubao Seed 2.1 Pro","description":"Higher-capability model in the Doubao Seed 2.1 family for agentic coding, long-context analysis, complex instruction following, and productivity workflows. Supports a 256k context window and up to 128k output tokens. Note: privacy and logging guarantees may be limited.","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":1,"output":5,"cache_read":0.5},"canonical_model_id":"bytedance-seed/seed-2.1-pro"},"bytedance/doubao-seed-character":{"id":"bytedance/doubao-seed-character","name":"Doubao Seed Character","description":"ByteDance's character-focused Doubao Seed model for roleplay, persona consistency, dialogue, and creative character interactions. It supports text and image input with a 128k context window. Requests route through ZenMux to ByteDance; ZenMux does not publish a model-API zero-retention or training guarantee, so avoid sensitive data.","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":0.15,"output":0.37,"cache_read":0.03,"cache_write":0.003},"canonical_model_id":"bytedance-seed/seed-character"},"stepfun-ai/step-3.5-flash-2603":{"id":"stepfun-ai/step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"stepfun-ai/step-3.5-flash":{"id":"stepfun-ai/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"stepfun/step-3.5-flash"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":1.2,"output":5.5,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.5:thinking":{"id":"moonshotai/kimi-k2.5:thinking","name":"Kimi K2.5 Thinking","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.375,"output":2,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.375,"output":2,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2-instruct":{"id":"moonshotai/kimi-k2-instruct","name":"Kimi K2 Instruct","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":8192},"cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/Kimi-K2-Instruct-0905":{"id":"moonshotai/Kimi-K2-Instruct-0905","name":"Kimi K2 0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":100352},"cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.44935,"output":1.892,"cache_read":0.07568},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-instruct-0711":{"id":"moonshotai/kimi-k2-instruct-0711","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/kimi-k2.6:thinking":{"id":"moonshotai/kimi-k2.6:thinking","name":"Kimi K2.6 Thinking","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.44935,"output":1.892,"cache_read":0.07568},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-latest":{"id":"moonshotai/kimi-latest","name":"Kimi Latest","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":1.2,"output":5.5,"cache_read":0.2}},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code High-Speed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":1.9,"output":8,"cache_read":0.32},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Meta's Muse Spark 1.3 Contributor is a frontier multimodal reasoning model for long-horizon coding and agentic workflows, with strong gains in computer use, browsing, professional tool use, codebase understanding, and million-token retrieval. It accepts text, images, audio, video, and files, supports tool calling and structured output, and always reasons before answering. Prompts and outputs may be used by Meta for training and to improve its products.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor (Data Used for Training)","description":"A much cheaper opt-in version of Muse Spark 1.2 with the same multimodal coding and agentic capabilities. Prompts and outputs sent to this Contributor model may be used by Meta for training and to improve its products; use the standard Muse Spark 1.2 model if you do not want your data used for training.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 2508","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.15},"canonical_model_id":"mistral/codestral-2508"},"mistralai/devstral-small-2505":{"id":"mistralai/devstral-small-2505","name":"Mistral Devstral Small 2505","description":"OpenHands+Devstral is 100% local 100% open, and is SOTA for the category on SWE-Bench Verified: 46.8% accuracy.","family":"devstral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-08-02","last_updated":"2025-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.06,"output":0.06,"cache_read":0.03}},"mistralai/mistral-nemo-instruct-2407":{"id":"mistralai/mistral-nemo-instruct-2407","name":"Mistral Nemo","description":"12B parameter model with multilingual support.","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.1003,"output":0.1207,"cache_read":0.05015},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-small-4-119b-2603:thinking":{"id":"mistralai/mistral-small-4-119b-2603:thinking","name":"Mistral Small 4 119B Thinking","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.4,"output":1.4,"cache_read":0.2},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/ministral-14b-instruct-2512":{"id":"mistralai/ministral-14b-instruct-2512","name":"Ministral 3 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Mistral Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":26214},"cost":{"input":0.1989,"output":0.595,"cache_read":0.09945},"canonical_model_id":"mistral/mistral-saba"},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.2,"output":0.2,"cache_read":0.1},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/mistral-medium-3.5:thinking":{"id":"mistralai/mistral-medium-3.5:thinking","name":"Mistral Medium 3.5 Thinking","description":"Mistral Medium 3.5 with reasoning enabled by default (reasoning_effort=high), for complex coding, agentic, and multi-step reasoning prompts.","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.5,"output":7.5,"cache_read":0.75}},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.4,"output":2,"cache_read":0.2},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.1,"output":0.1,"cache_read":0.05},"canonical_model_id":"mistral/ministral-3b-2512"},"mistralai/devstral-2-123b-instruct-2512":{"id":"mistralai/devstral-2-123b-instruct-2512","name":"Devstral 2 123B","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.4,"output":1.4,"cache_read":0.2},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Mistral Large 2411","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":102400},"cost":{"input":2.006,"output":6.001,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-small-4-119b-2603":{"id":"mistralai/mistral-small-4-119b-2603","name":"Mistral Small 4 119B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.4,"output":1.4,"cache_read":0.2},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mixtral-8x22b-instruct-v0.1":{"id":"mistralai/mixtral-8x22b-instruct-v0.1","name":"Mixtral 8x22B","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":52428},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Mistral Small 3.2 24B (2506)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.2,"output":0.4,"cache_read":0.1},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-medium-3.5":{"id":"mistralai/mistral-medium-3.5","name":"Mistral Medium 3.5","description":"Mistral Medium 3.5 is a 128B dense open-weights flagship model for instruction-following, reasoning, coding, long-horizon agentic work, tool use, structured output, and multimodal prompts. It supports a 256k context window and configurable reasoning effort.","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.5,"output":7.5,"cache_read":0.75}},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.4,"output":2,"cache_read":0.2},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/mistral-large-4":{"id":"mistralai/mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Mistral Small 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.1155,"output":0.3465},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B (2503)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":102400},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 8B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.15,"output":0.15,"cache_read":0.075},"canonical_model_id":"mistral/ministral-8b-2512"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1001536,"input":1001536,"output":32768},"cost":{"input":0.15,"output":0.7,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.6-27b:thinking":{"id":"qwen/qwen3.6-27b:thinking","name":"Qwen3.6 27B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.203,"output":2.24,"cache_read":0.1015},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3-vl-235b-a22b-thinking":{"id":"qwen/qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"qwen/qwen3.8-max-prime":{"id":"qwen/qwen3.8-max-prime","name":"Qwen3.8 Max Prime","description":"High-throughput edition of Qwen3.8 Max for coding, professional work, multimodal understanding, and long-running agent workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"qwen/qwen3.8-27b:thinking":{"id":"qwen/qwen3.8-27b:thinking","name":"Qwen3.8 27B Thinking","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1001536,"input":1001536,"output":32768},"cost":{"input":0.15,"output":0.7,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3-max-2026-01-23":{"id":"qwen/qwen3-max-2026-01-23","name":"Qwen3 Max 2026-01-23","description":"Qwen3 Max is Alibaba's flagship Qwen 3 reasoning model with native tool use (web search, web extractor, code interpreter) and a 256K context window.","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-01-26","last_updated":"2026-01-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.2002,"output":6.001,"cache_read":0.6001}},"qwen/qwen3.8-27b-fable":{"id":"qwen/qwen3.8-27b-fable","name":"Qwen 3.8 27B Fable","description":"Qwen 3.8 27B Fable is an open-weight multimodal creative finetune for expressive dialogue, long-form storytelling, character work, and roleplay.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.5,"cache_read":0.125}},"qwen/qwen3.8-27b-obliterated":{"id":"qwen/qwen3.8-27b-obliterated","name":"Qwen 3.8 27B Obliterated","description":"Qwen 3.8 27B Obliterated is an open-weight multimodal model LoRA-tuned for fewer refusals across chat, coding, reasoning, tool use, and long-context work.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.75,"cache_read":0.225}},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.3,"output":1.9,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen 3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.325,"output":1.95,"cache_read":0.0325,"cache_write":0.40625},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-235b-a22b":{"id":"qwen/qwen3-235b-a22b","name":"Qwen 3 235b A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.0875,"output":0.35,"cache_read":0.0175},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"qwen/qwen3.7-plus:thinking":{"id":"qwen/qwen3.7-plus:thinking","name":"Qwen3.7 Plus Thinking","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.8-27b-obliterated:thinking":{"id":"qwen/qwen3.8-27b-obliterated:thinking","name":"Qwen 3.8 27B Obliterated Thinking","description":"Qwen 3.8 27B Obliterated with thinking enabled for more deliberate creative work, coding, multimodal analysis, tool use, and long-context problem solving.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.75,"cache_read":0.225}},"qwen/qwen3-coder-30b-a3b-instruct":{"id":"qwen/qwen3-coder-30b-a3b-instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen 3 32b","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3.8-27b-antislop":{"id":"qwen/qwen3.8-27b-antislop","name":"Qwen 3.8 27B Antislop","description":"Qwen 3.8 27B Antislop is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3.5-plus:thinking":{"id":"qwen/qwen3.5-plus:thinking","name":"Qwen3.5 Plus Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.4,"output":2.4,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.5-omni-flash":{"id":"qwen/qwen3.5-omni-flash","name":"Qwen3.5 Omni Flash","description":"Qwen3.5 Omni Flash is Qwen's fast multimodal model. We verified live support for text prompts, images, audio files, and direct video URLs on Alibaba's chat-completions-compatible API. Alibaba describes Flash as a fully evolved version of Qwen3 Omni with audio input support across 60+ languages.","family":"qwen3.5","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-03-30","last_updated":"2026-03-30","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":49152,"input":49152,"output":16384}},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.7-flash:thinking":{"id":"qwen/qwen3.7-flash:thinking","name":"Qwen3.7 Flash Thinking","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.15,"output":0.65,"cache_read":0.075},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3.8-27b-kiwen":{"id":"qwen/qwen3.8-27b-kiwen","name":"Qwen 3.8 27B Kiwen","description":"Qwen 3.8 27B Kiwen is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.5-35b-a3b:thinking":{"id":"qwen/qwen3.5-35b-a3b:thinking","name":"Qwen3.5 35B A3B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-8b":{"id":"qwen/qwen3-8b","name":"Qwen 3 8B","description":"Dense 8B Qwen3 model with switchable thinking for chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.47,"output":0.47,"cache_read":0.235},"canonical_model_id":"alibaba/qwen3-8b"},"qwen/qwen3-235b-a22b-instruct-2507":{"id":"qwen/qwen3-235b-a22b-instruct-2507","name":"Qwen 3 235b A22B 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.0875,"output":0.35,"cache_read":0.0175},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.8-27b-uncensored":{"id":"qwen/qwen3.8-27b-uncensored","name":"Qwen 3.8 27B Uncensored","description":"Qwen 3.8 27B Uncensored is an NVFP4 open-weight multimodal model LoRA-tuned for fewer refusals across chat, coding, reasoning, tool use, and long-context work.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"input":1010000,"output":32768},"cost":{"input":1,"output":3,"cache_read":0.22}},"qwen/qwen3.5-flash:thinking":{"id":"qwen/qwen3.5-flash:thinking","name":"Qwen3.5 Flash Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.8-27b-cybersecurity":{"id":"qwen/qwen3.8-27b-cybersecurity","name":"Qwen 3.8 27B Cybersecurity","description":"Qwen 3.8 27B Cybersecurity is a cybersecurity-focused variant based on the uncensored model, with provider moderation for illegal activities. It supports optional reasoning, image understanding, tool calling, and a 262,144-token context window.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-19","last_updated":"2026-09-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.2002,"output":6.001,"cache_read":0.6001},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.05,"output":0.15,"cache_read":0.025},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen 3 235b A22B 2507 Thinking","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":81920},"cost":{"input":0.3,"output":0.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.26,"output":2.08,"cache_read":0.103788},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen2.5-coder-32b-instruct":{"id":"qwen/qwen2.5-coder-32b-instruct","name":"Qwen 2.5 Coder 32b","description":"Open coding-focused Qwen model for code generation, repair, and repository reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"qwen/qwen-long":{"id":"qwen/qwen-long","name":"Qwen Long 10M","description":"Alibaba's huge context window model. Takes in up to 10 million tokens, which is equivalent to dozens of books.","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"input":10000000,"output":8192},"cost":{"input":0.1003,"output":0.408,"cache_read":0.05015}},"qwen/qwen3.8-27b-kimiko":{"id":"qwen/qwen3.8-27b-kimiko","name":"Qwen 3.8 27B Kimiko","description":"Qwen 3.8 27B Kimiko is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.6-max-preview":{"id":"qwen/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":245760,"input":245760,"output":65536},"cost":{"input":1.04,"output":6.24,"cache_read":0.52},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"qwen/qwen-turbo":{"id":"qwen/qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":8192},"cost":{"input":0.04998,"output":0.2006,"cache_read":0.02499},"canonical_model_id":"alibaba/qwen-turbo"},"qwen/qwen3.8-max:thinking":{"id":"qwen/qwen3.8-max:thinking","name":"Qwen3.8 Max Thinking","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"input":991000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.2,"output":0.5,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.39,"output":2.34,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen-2.5-72b-instruct":{"id":"qwen/qwen-2.5-72b-instruct","name":"Qwen2.5 72B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-07-03","last_updated":"2025-07-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":8192},"cost":{"input":0.36,"output":0.4,"cache_read":0.1785}},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"input":991000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-27b-swift":{"id":"qwen/qwen3.8-27b-swift","name":"Qwen 3.8 27B Swift","description":"Qwen 3.8 27B Swift is a finetune for fast responses, served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.6-35b-a3b:thinking":{"id":"qwen/qwen3.6-35b-a3b:thinking","name":"Qwen3.6 35B A3B Thinking","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.112,"output":0.8,"cache_read":0.056},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.8-27b-uncensored:thinking":{"id":"qwen/qwen3.8-27b-uncensored:thinking","name":"Qwen 3.8 27B Uncensored Thinking","description":"Qwen 3.8 27B Uncensored with thinking enabled for more deliberate creative work, coding, multimodal analysis, tool use, and long-context problem solving.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"input":1010000,"output":32768},"cost":{"input":1,"output":3,"cache_read":0.22}},"qwen/qwen3-14b":{"id":"qwen/qwen3-14b","name":"Qwen 3 14b","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.08,"output":0.24,"cache_read":0.04}},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.4,"output":2.4,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.7-max:thinking":{"id":"qwen/qwen3.7-max:thinking","name":"Qwen3.7 Max Thinking","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.5-27b:thinking":{"id":"qwen/qwen3.5-27b:thinking","name":"Qwen3.5 27B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.135},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.8-27b-hemmingway":{"id":"qwen/qwen3.8-27b-hemmingway","name":"Qwen 3.8 27B Hemingway","description":"Qwen 3.8 27B Hemingway is an open-weight NVFP4 multimodal creative finetune for long-form prose, character dialogue, storytelling, and roleplay.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.25,"output":1.5,"cache_read":0.125}},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.203,"output":2.24,"cache_read":0.1015},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":131072},"cost":{"input":0.14,"output":0.42,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":995904,"input":995904,"output":32768},"cost":{"input":0.3995,"output":1.2002,"cache_read":0.19975},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3.8-27b-novelist":{"id":"qwen/qwen3.8-27b-novelist","name":"Qwen 3.8 27B Novelist","description":"Qwen 3.8 27B Novelist is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen-max":{"id":"qwen/qwen-max","name":"Qwen 2.5 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":1.5997,"output":6.392,"cache_read":0.79985},"canonical_model_id":"alibaba/qwen-max"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.8-27b-darkidol":{"id":"qwen/qwen3.8-27b-darkidol","name":"Qwen 3.8 27B DarkIdol","description":"Qwen 3.8 27B DarkIdol is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.8-27b-fimi":{"id":"qwen/qwen3.8-27b-fimi","name":"Qwen 3.8 27B Fimi","description":"Qwen 3.8 27B Fimi is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.135},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.5-omni-plus":{"id":"qwen/qwen3.5-omni-plus","name":"Qwen3.5 Omni Plus","description":"Qwen3.5 Omni Plus is Qwen's stronger general multimodal model. We verified live support for text prompts, images, audio files, and direct video URLs on Alibaba's chat-completions-compatible API. Alibaba describes Plus as a comprehensive evolution of Qwen3 Omni with support for over 10 hours of audio input.","family":"qwen3.5","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-03-30","last_updated":"2026-03-30","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536}},"qwen/qwen3.8-27b-queen":{"id":"qwen/qwen3.8-27b-queen","name":"Qwen 3.8 27B Queen","description":"Qwen 3.8 27B Queen is an open-weight roleplay finetune with image understanding, tool calling, optional reasoning, and a 262,144-token context window.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.5,"cache_read":0.125}},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.250251,"output":0.80293,"cache_read":0.079764},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B (Instruct)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":1.1,"cache_read":0.07},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3.5-397b-a17b:thinking":{"id":"qwen/qwen3.5-397b-a17b:thinking","name":"Qwen3.5 397B A17B Thinking","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.39,"output":2.34,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.112,"output":0.8,"cache_read":0.056},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.8-27b-synthia":{"id":"qwen/qwen3.8-27b-synthia","name":"Qwen 3.8 27B Synthia","description":"Qwen 3.8 27B Synthia is a text-only roleplay finetune served by Aoru in NVFP4 with a 262,144-token context window.","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.5-122b-a10b:thinking":{"id":"qwen/qwen3.5-122b-a10b:thinking","name":"Qwen3.5 122B A10B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.26,"output":2.08,"cache_read":0.103788},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.19,"output":1.16,"cache_read":0.02,"cache_write":0.24},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3-coder":{"id":"qwen/qwen3-coder","name":"Qwen 3 Coder 480B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"input":262000,"output":65536},"cost":{"input":0.13,"output":0.5,"cache_read":0.065},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"elorian/glimpse":{"id":"elorian/glimpse","name":"Elorian Glimpse","description":"Glimpse is Elorian's visual reasoning model for understanding images and video and answering in text. It can analyze charts, count objects, follow events over time, and reason about spatial relationships and physical cause and effect. It supports tool calling and structured output across a shared 262,144-token context window, with up to 256,000 output tokens within the remaining context.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":256000},"cost":{"input":0.63,"output":3.15,"cache_read":0.063}},"soob3123/Veiled-Calla-12B":{"id":"soob3123/Veiled-Calla-12B","name":"Veiled Calla 12B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-04-13","last_updated":"2025-04-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"soob3123/amoral-gemma3-27B-v2":{"id":"soob3123/amoral-gemma3-27B-v2","name":"Amoral Gemma3 27B v2","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-05-23","last_updated":"2025-05-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"liquid/lfm-2.5-24b-a2b":{"id":"liquid/lfm-2.5-24b-a2b","name":"LFM2.5 24B A2B","description":"Liquid AI's text-only mixture-of-experts model for reasoning and tool-using agents, with 24B total and 2.45B active parameters. It always thinks before answering and supports structured output. Pair it with retrieval for knowledge-heavy tasks. Open weights use the LFM Open License v1.0, with a $10M annual-revenue commercial threshold; hosted API use is subject to separate service terms that permit retention and use of prompts and responses for model training.","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-14","last_updated":"2026-10-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"liquid/lfm-2.5-2.6b":{"id":"liquid/lfm-2.5-2.6b","name":"LFM2.5 2.6B","description":"Liquid AI's compact 2.6B reasoning model for agent workflows, data extraction, RAG, and long-context processing. It supports tool calling and structured output, but Liquid advises against using it for agentic coding. Warning: prompts and responses may be logged and used for model training or service improvement; do not send sensitive data.","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":0.1,"output":0.2,"cache_read":0.05}},"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1":{"id":"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1","name":"EVA-LLaMA-3.33-70B-v0.1","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":2.006,"output":2.006,"cache_read":1.003}},"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0":{"id":"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0","name":"EVA Llama 3.33 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":2.006,"output":2.006,"cache_read":1.003}},"EVA-UNIT-01/EVA-Qwen2.5-72B-v0.2":{"id":"EVA-UNIT-01/EVA-Qwen2.5-72B-v0.2","name":"EVA-Qwen2.5-72B-v0.2","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.799,"output":0.799,"cache_read":0.3995}},"EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2":{"id":"EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2","name":"EVA-Qwen2.5-32B-v0.2","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.799,"output":0.799,"cache_read":0.3995}},"NousResearch/hermes-3-llama-3.1-70b":{"id":"NousResearch/hermes-3-llama-3.1-70b","name":"Hermes 3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.408,"output":0.408,"cache_read":0.204},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"NousResearch/hermes-4-405b:thinking":{"id":"NousResearch/hermes-4-405b:thinking","name":"Hermes 4 Large (Thinking)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.15},"canonical_model_id":"nousresearch/hermes-4-405b"},"NousResearch/hermes-4-405b":{"id":"NousResearch/hermes-4-405b","name":"Hermes 4 Large","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.15},"canonical_model_id":"nousresearch/hermes-4-405b"},"tencent/hy3":{"id":"tencent/hy3","name":"Tencent Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":128000},"cost":{"input":0.066,"output":0.26,"cache_read":0.029},"canonical_model_id":"tencent/hy3"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Tencent Hy4 Preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":64000},"cost":{"input":0.7506,"output":2.2509,"cache_read":0.0378},"canonical_model_id":"tencent/hy4-preview"},"TheDrummer/skyfall-31b-v4.2":{"id":"TheDrummer/skyfall-31b-v4.2","name":"TheDrummer Skyfall 31B v4.2","description":"TheDrummer's Skyfall 31B v4.2 is a Magistral Small finetune for creative writing, roleplay, and dialogue. This text-only deployment uses NVFP4 weights and supports a 16K-token context window.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":4096},"cost":{"input":0.12,"output":0.15,"cache_read":0.06}},"TheDrummer/Cydonia-24B-v4":{"id":"TheDrummer/Cydonia-24B-v4","name":"The Drummer Cydonia 24B v4","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.2006,"output":0.2414,"cache_read":0.1003}},"TheDrummer/Anubis-70B-v1.1":{"id":"TheDrummer/Anubis-70B-v1.1","name":"Anubis 70B v1.1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":16384},"cost":{"input":0.31,"output":0.31,"cache_read":0.155}},"TheDrummer/Cydonia-24B-v4.3":{"id":"TheDrummer/Cydonia-24B-v4.3","name":"The Drummer Cydonia 24B v4.3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-25","last_updated":"2025-12-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.12,"output":0.15,"cache_read":0.06}},"TheDrummer/Cydonia-24B-v4.1":{"id":"TheDrummer/Cydonia-24B-v4.1","name":"The Drummer Cydonia 24B v4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.3,"output":0.5,"cache_read":0.15},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"TheDrummer/orion-26b-a4b-v1.1":{"id":"TheDrummer/orion-26b-a4b-v1.1","name":"Orion 26B A4B v1.1","description":"TheDrummer's Gemma 4 26B A4B fine-tune for creative writing, roleplay, and expressive dialogue. Supports optional thinking with a 65,536-token context window and up to 8,192 output tokens. Text input only.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"TheDrummer/Artemis-v1.1":{"id":"TheDrummer/Artemis-v1.1","name":"TheDrummer/Artemis v1.1","description":"TheDrummer's Artemis v1.1 is a Gemma 4 31B fine-tune for creative writing, expressive dialogue, and roleplay, with optional thinking and a 262K context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-06","last_updated":"2026-09-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"TheDrummer/Rocinante-12B-v1.1":{"id":"TheDrummer/Rocinante-12B-v1.1","name":"Rocinante 12b","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.408,"output":0.595,"cache_read":0.204}},"TheDrummer/skyfall-36b-v2":{"id":"TheDrummer/skyfall-36b-v2","name":"TheDrummer Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":29491},"cost":{"input":0.55,"output":0.8,"cache_read":0.25},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"TheDrummer/Magidonia-24B-v4.3":{"id":"TheDrummer/Magidonia-24B-v4.3","name":"The Drummer Magidonia 24B v4.3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-25","last_updated":"2025-12-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.1003,"output":0.1207,"cache_read":0.05015}},"TheDrummer/Anubis-70B-v1":{"id":"TheDrummer/Anubis-70B-v1","name":"Anubis 70B v1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":16384},"cost":{"input":0.31,"output":0.31,"cache_read":0.155}},"TheDrummer/Cydonia-24B-v2":{"id":"TheDrummer/Cydonia-24B-v2","name":"The Drummer Cydonia 24B v2","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.1003,"output":0.1207,"cache_read":0.05015}},"TheDrummer/UnslopNemo-12B-v4.1":{"id":"TheDrummer/UnslopNemo-12B-v4.1","name":"UnslopNemo 12b v4","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"input":8192,"output":26214},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.015},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.025},"canonical_model_id":"inference-net/schematron-v2-small"},"ornith-ai/ornith-1.5-122b":{"id":"ornith-ai/ornith-1.5-122b","name":"Ornith 1.5 122B","description":"Ornith 1.5 122B is an MIT-licensed open-weight model for coding, reasoning, tool use and image understanding. It accepts text and images across a shared 262,144-token context, with up to 32,768 output tokens. Reasoning is enabled by default; JSON object output and streaming are supported.","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.2,"output":0.8,"cache_read":0.05}},"ornith-ai/ornith-1.5-35b-a3b:thinking":{"id":"ornith-ai/ornith-1.5-35b-a3b:thinking","name":"Ornith 1.5 35B Thinking","description":"Ornith 1.5 35B A3B is an open-weight mixture-of-experts model for agentic coding, reasoning, tool use, image understanding, and long-context work. This variant enables thinking by default.","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.05}},"ornith-ai/ornith-1.5-35b-a3b":{"id":"ornith-ai/ornith-1.5-35b-a3b","name":"Ornith 1.5 35B","description":"Ornith 1.5 35B A3B is an open-weight mixture-of-experts model for agentic coding, tool use, image understanding, and long-context work. This variant disables thinking for faster direct responses.","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.05}},"anthropic/claude-opus-4.1:thinking:32768":{"id":"anthropic/claude-opus-4.1:thinking:32768","name":"Claude 4.1 Opus Thinking (32K)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-opus-4.1:thinking:1024":{"id":"anthropic/claude-opus-4.1:thinking:1024","name":"Claude 4.1 Opus Thinking (1K)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude 4.6 Opus","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-4.5:thinking":{"id":"anthropic/claude-sonnet-4.5:thinking","name":"Claude Sonnet 4.5 Thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude 4.5 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4.7:thinking":{"id":"anthropic/claude-opus-4.7:thinking","name":"Claude 4.7 Opus Thinking","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-haiku-4.5:thinking":{"id":"anthropic/claude-haiku-4.5:thinking","name":"Claude Haiku 4.5 Thinking","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-opus-4.6:thinking:max":{"id":"anthropic/claude-opus-4.6:thinking:max","name":"Claude 4.6 Opus Thinking Max","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.6:thinking":{"id":"anthropic/claude-opus-4.6:thinking","name":"Claude 4.6 Opus Thinking","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5:thinking":{"id":"anthropic/claude-opus-4.5:thinking","name":"Claude 4.5 Opus Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude 4.1 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-opus-4.1:thinking:8192":{"id":"anthropic/claude-opus-4.1:thinking:8192","name":"Claude 4.1 Opus Thinking (8K)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-opus-4.6:thinking:low":{"id":"anthropic/claude-opus-4.6:thinking:low","name":"Claude 4.6 Opus Thinking Low","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-4.6:thinking":{"id":"anthropic/claude-sonnet-4.6:thinking","name":"Claude Sonnet 4.6 Thinking","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-sonnet-5.5:thinking":{"id":"anthropic/claude-sonnet-5.5:thinking","name":"Claude Sonnet 5.5 Thinking","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude 4.7 Opus","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-latest":{"id":"anthropic/claude-fable-latest","name":"Claude Fable Latest","description":"Compatibility alias for Claude Fable.","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5}},"anthropic/claude-opus-4.6:thinking:medium":{"id":"anthropic/claude-opus-4.6:thinking:medium","name":"Claude 4.6 Opus Thinking Medium","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-5:thinking":{"id":"anthropic/claude-sonnet-5:thinking","name":"Claude Sonnet 5 Thinking","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-sonnet-latest":{"id":"anthropic/claude-sonnet-latest","name":"Claude Sonnet Latest","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-01","last_updated":"2026-03-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4.8:thinking":{"id":"anthropic/claude-opus-4.8:thinking","name":"Claude Opus 4.8 Thinking","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-haiku-latest":{"id":"anthropic/claude-haiku-latest","name":"Claude Haiku Latest","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]}},"anthropic/claude-opus-latest":{"id":"anthropic/claude-opus-latest","name":"Claude Opus Latest","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5}},"anthropic/claude-opus-4.1:thinking":{"id":"anthropic/claude-opus-4.1:thinking","name":"Claude 4.1 Opus Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.7-flash:thinking":{"id":"z-ai/glm-4.7-flash:thinking","name":"GLM 4.7 Flash Thinking","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"input":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/GLM-4.5:thinking":{"id":"z-ai/GLM-4.5:thinking","name":"GLM 4.5 (Thinking)","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.3,"output":1.3,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5.3-flash-apex-abliterated":{"id":"z-ai/glm-5.3-flash-apex-abliterated","name":"Apex Flash 1 Abliterated","description":"Apex Flash 1 Abliterated is an experimental derivative of Cantina Security and Yeta's Apex Flash 1 with modified refusal behavior.","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.5,"cache_read":0.075}},"z-ai/GLM-4.6-turbo":{"id":"z-ai/GLM-4.6-turbo","name":"GLM 4.6 Turbo","description":"Fast variant of GLM 4.6 for general chat, coding, and analysis with improved latency and strong reasoning.","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":1,"output":3,"cache_read":0.5}},"z-ai/glm-5-original:thinking":{"id":"z-ai/glm-5-original:thinking","name":"GLM 5 Original Thinking","description":"GLM-5 original with extended thinking capabilities for complex reasoning.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.6-original":{"id":"z-ai/glm-4.6-original","name":"GLM 4.6 Original","description":"GLM-4.6, Zhipu's flagship text model with 256K context window and advanced reasoning capabilities. Direct via Z-AI (Zhipu).","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":65535},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.6v-original":{"id":"z-ai/glm-4.6v-original","name":"GLM 4.6V Original","description":"GLM-4.6V scales its context window to 128k tokens in training, and achieves SoTA performance in visual understanding among models of similar parameter scales. Integrates native Function Calling capabilities, bridging 'visual perception' and 'executable action' for multimodal agents. Direct via Z-AI (Zhipu).","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":24000},"cost":{"input":0.6,"output":0.9,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-5-original":{"id":"z-ai/glm-5-original","name":"GLM 5 Original","description":"GLM-5 is Zhipu's latest flagship model with advanced reasoning and instruction following. Routed directly via Z-AI (Zhipu).","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-5.2:thinking":{"id":"z-ai/glm-5.2:thinking","name":"GLM 5.2 Thinking","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.189,"output":0.594,"cache_read":0.0351},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-latest":{"id":"z-ai/glm-latest","name":"GLM Latest","description":"Compatibility alias that routes to the newest thinking GLM model. Currently routes to GLM 5.2 Thinking.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"z-ai/glm-4.7-original":{"id":"z-ai/glm-4.7-original","name":"GLM 4.7 Original","description":"GLM-4.7 is a next-gen GLM series text model with stronger reasoning, long-context chat, and reliable tool use. Routed directly via Z-AI (Zhipu).","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.4004,"output":1.9292,"cache_read":0.08008},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.3,"output":1.3,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-4.6:thinking":{"id":"z-ai/glm-4.6:thinking","name":"GLM 4.6 Thinking","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-5.1:thinking":{"id":"z-ai/glm-5.1:thinking","name":"GLM 5.1 Thinking","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5.3:thinking":{"id":"z-ai/glm-5.3:thinking","name":"GLM 5.3 Thinking","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-uncensored":{"id":"z-ai/glm-5.3-uncensored","name":"GLM 5.3 Uncensored","description":"An experimental uncensored version of the full GLM 5.3 reasoning model for chat, creative writing, coding, and tool use. It may show unusual behavior or inconsistent response quality.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-26","last_updated":"2026-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.25,"output":2.25,"cache_read":0.3}},"z-ai/glm-4.5v:thinking":{"id":"z-ai/glm-4.5v:thinking","name":"GLM 4.5V Thinking","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-5.3-flash-apex":{"id":"z-ai/glm-5.3-flash-apex","name":"Apex Flash 1","description":"Apex Flash 1 is Cantina Security and Yeta's cybersecurity-focused post-train of GLM 5.3 Flash for code investigation and tool use.","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.5,"cache_read":0.075}},"z-ai/glm-5v-turbo:thinking":{"id":"z-ai/glm-5v-turbo:thinking","name":"GLM 5V Turbo Thinking","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"input":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.189,"output":0.594,"cache_read":0.0351},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5:thinking":{"id":"z-ai/glm-5:thinking","name":"GLM 5 Thinking","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"input":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM 4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":24000},"cost":{"input":0.3,"output":0.9,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/GLM-4.5-Air":{"id":"z-ai/GLM-4.5-Air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":98304},"cost":{"input":0.12,"output":0.8,"cache_read":0.06},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-5.3-flash-uncensored":{"id":"z-ai/glm-5.3-flash-uncensored","name":"GLM 5.3 Flash Uncensored","description":"GLM 5.3 Flash Uncensored is an uncensored fine-tune of the efficient 320B mixture-of-experts reasoning model, built for unrestricted chat, creative writing, coding, agentic work, tool use, and long-context tasks.","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.2,"output":0.8,"cache_read":0.07}},"z-ai/glm-4.7:thinking":{"id":"z-ai/glm-4.7:thinking","name":"GLM 4.7 Thinking","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.4004,"output":1.9292,"cache_read":0.08008},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.3-flash-cybersecurity":{"id":"z-ai/glm-5.3-flash-cybersecurity","name":"GLM 5.3 Flash Cybersecurity","description":"GLM 5.3 Flash Cybersecurity is a cybersecurity-focused variant based on the uncensored model, with provider moderation for illegal activities. It supports always-on reasoning, image understanding, tool calling, and a 1,048,576-token context window.","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-19","last_updated":"2026-09-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.5,"cache_read":0.075}},"z-ai/glm-4.7-original:thinking":{"id":"z-ai/glm-4.7-original:thinking","name":"GLM 4.7 Original Thinking","description":"GLM-4.7 original with extended thinking capabilities for complex reasoning.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/GLM-4.6-turbo:thinking":{"id":"z-ai/GLM-4.6-turbo:thinking","name":"GLM 4.6 Turbo (Thinking)","description":"GLM 4.6 Turbo with thinking mode enabled for enhanced reasoning; shows internal reasoning and supports long context.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":1,"output":3,"cache_read":0.5}},"z-ai/GLM-4.5-Air:thinking":{"id":"z-ai/GLM-4.5-Air:thinking","name":"GLM 4.5 Air (Thinking)","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":98304},"cost":{"input":0.12,"output":0.8,"cache_read":0.06},"canonical_model_id":"zhipuai/glm-4.5-air"},"anthracite-org/magnum-v2-72b":{"id":"anthracite-org/magnum-v2-72b","name":"Magnum V2 72B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":2.006,"output":2.992,"cache_read":1.003}},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":2.006,"output":2.992,"cache_read":1.003},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"meta-llama/llama-4-scout":{"id":"meta-llama/llama-4-scout","name":"Llama 4 Scout","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":328000,"input":328000,"output":65536},"cost":{"input":0.085,"output":0.46,"cache_read":0.0425},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama 3.3 70b Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.05,"output":0.23,"cache_read":0.025},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama 3.1 8b Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.02,"output":0.04,"cache_read":0.0272},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-3.2-3b-instruct":{"id":"meta-llama/llama-3.2-3b-instruct","name":"Llama 3.2 3b Instruct","description":"Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":8192},"cost":{"input":0.0306,"output":0.0493,"cache_read":0.0153},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"meta-llama/llama-4-maverick":{"id":"meta-llama/llama-4-maverick","name":"Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.1875,"output":0.6525,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"nvidia/nemotron-3-ultra-550b-a55b:thinking":{"id":"nvidia/nemotron-3-ultra-550b-a55b:thinking","name":"Nvidia Nemotron 3 Ultra 550B Thinking","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nvidia Nemotron 3 Nano 30B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":235929},"cost":{"input":0.17,"output":0.68,"cache_read":0.085},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"Nvidia Nemotron 3.5 Super VL","description":"NVIDIA’s open-weight 120B model (12B active) for reasoning, coding, tool use, document and GUI understanding, and video understanding. Text, images and video in; text out. Served in NVFP4 with a 262,144-token shared context window. NanoGPT currently limits responses to 16,384 tokens while separate hosted output limits remain unpublished. OpenMDW 1.1 license.","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nvidia Nemotron 3.5 Lightning","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nvidia Nemotron 3 Ultra 550B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-super-120b-a12b:thinking":{"id":"nvidia/nemotron-3-super-120b-a12b:thinking","name":"Nvidia Nemotron 3 Super 120B Thinking","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.05,"output":0.25,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/Llama-3.1-Nemotron-70B-Instruct-HF":{"id":"nvidia/Llama-3.1-Nemotron-70B-Instruct-HF","name":"Nvidia Nemotron 70b","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.357,"output":0.408,"cache_read":0.1785},"canonical_model_id":"nvidia/llama-3.1-nemotron-70b-instruct"},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.15,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3.5-lightning:thinking":{"id":"nvidia/nemotron-3.5-lightning:thinking","name":"Nvidia Nemotron 3.5 Lightning Thinking","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nvidia Nemotron 3 Super 120B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.05,"output":0.25,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/Llama-3.3-Nemotron-Super-49B-v1":{"id":"nvidia/Llama-3.3-Nemotron-Super-49B-v1","name":"Nvidia Nemotron Super 49B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-18","last_updated":"2025-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.15,"output":0.15,"cache_read":0.075},"canonical_model_id":"nvidia/llama-3.3-nemotron-super-49b-v1"},"nothingiisreal/L3.1-70B-Celeste-V0.1-BF16":{"id":"nothingiisreal/L3.1-70B-Celeste-V0.1-BF16","name":"Llama 3.1 70B Celeste v0.1","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"unbiased/pareto-26.10":{"id":"unbiased/pareto-26.10","name":"Pareto 26.10","description":"Pareto 26.10 is Unbiased AI's multimodal composite model for research, coding, and agentic workflows. It accepts text and images, supports tool calling, and provides a 1M-token context window with up to 131K output tokens. This is the stable release of Pareto 26.10. Tool selection is automatic; forced tool selection, structured output, and adjustable reasoning controls are not supported.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"unbiased/pareto-26.9":{"id":"unbiased/pareto-26.9","name":"Pareto 26.9","description":"Pareto 26.9 is an experimental multimodal composite model from Unbiased AI for research, coding, agentic workflows, and general-purpose tasks. It runs several frontier and open models for each request and returns a single selected or synthesized answer. This experimental release supports text and image input, tool calling, a 262K-token context window, and up to 131K output tokens. Its underlying model mix and behavior may change as it is evaluated. Request content may be retained for abuse prevention and security monitoring for up to 30 days, and is not used for model training without explicit consent.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25}},"unsloth/gemma-3-27b-it":{"id":"unsloth/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":96000},"cost":{"input":0.2992,"output":0.2992,"cache_read":0.1496},"canonical_model_id":"google/gemma-3-27b-it"},"unsloth/gemma-3-12b-it":{"id":"unsloth/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.272,"output":0.272,"cache_read":0.136},"canonical_model_id":"google/gemma-3-12b-it"},"unsloth/gemma-3-4b-it":{"id":"unsloth/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003},"canonical_model_id":"google/gemma-3-4b-it"},"prism-ml/ternary-bonsai-2-27b":{"id":"prism-ml/ternary-bonsai-2-27b","name":"Ternary Bonsai 2 27B","description":"Ternary Bonsai 2 27B is a 27B-parameter reasoning model from PrismML derived from Qwen3.8-27B. It supports coding, mathematics, tool calling, and image understanding with a 262K-token context window.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.5,"cache_read":0.0375}},"prism-ml/ternary-bonsai-2-9b":{"id":"prism-ml/ternary-bonsai-2-9b","name":"Ternary Bonsai2 9B","description":"Ternary Bonsai2 9B is Prism ML’s open-weight model derived from Qwen3.5-9B with ternary language weights. It accepts text and generates text, supports reasoning and tool calling, and offers a shared 262,144-token context window with up to 65,536 output tokens. Thinking is enabled by default and can be turned off.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.08,"output":0.13,"cache_read":0.04}},"pamanseau/OpenReasoning-Nemotron-32B":{"id":"pamanseau/OpenReasoning-Nemotron-32B","name":"OpenReasoning Nemotron 32B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05}},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember 1","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"inflatebot/MN-12B-Mag-Mell-R1":{"id":"inflatebot/MN-12B-Mag-Mell-R1","name":"Mag Mell R1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"nbeerbower/gemma4-gutenberg-31b":{"id":"nbeerbower/gemma4-gutenberg-31b","name":"Gemma 4 Gutenberg 31B","description":"nbeerbower's Gemma 4 31B finetune for literary fiction, character interiority, and controlled narrative pacing. Its text decoder was trained with ORPO on 14,816 Athanorlite-DPO preference pairs. This text-only deployment supports a 65,536-token context window and up to 8,192 output tokens.","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"mlabonne/NeuralDaredevil-8B-abliterated":{"id":"mlabonne/NeuralDaredevil-8B-abliterated","name":"Neural Daredevil 8B abliterated","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":0.44,"output":0.44,"cache_read":0.22}},"Steelskull/L3.3-Cu-Mai-R1-70b":{"id":"Steelskull/L3.3-Cu-Mai-R1-70b","name":"Llama 3.3 70B Cu Mai","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Steelskull/L3.3-MS-Nevoria-70b":{"id":"Steelskull/L3.3-MS-Nevoria-70b","name":"Steelskull Nevoria 70b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Steelskull/L3.3-Electra-R1-70b":{"id":"Steelskull/L3.3-Electra-R1-70b","name":"Steelskull Electra R1 70b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.69989,"output":0.69989,"cache_read":0.349945}},"Steelskull/L3.3-Nevoria-R1-70b":{"id":"Steelskull/L3.3-Nevoria-R1-70b","name":"Steelskull Nevoria R1 70b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"TEE/qwen3.8-27b":{"id":"TEE/qwen3.8-27b","name":"Qwen3.8 27B TEE","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.4,"output":3,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.8-27b"},"TEE/qwen2.5-vl-72b-instruct":{"id":"TEE/qwen2.5-vl-72b-instruct","name":"Qwen2.5 VL 72B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"TEE/deepseek-v4.1-flash":{"id":"TEE/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash TEE","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.65,"output":1.45,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"TEE/gemma4-31b:thinking":{"id":"TEE/gemma4-31b:thinking","name":"Gemma 4 31B Thinking TEE","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.4,"output":1,"cache_read":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"TEE/glm-5.3":{"id":"TEE/glm-5.3","name":"GLM 5.3 TEE","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"TEE/kimi-k3":{"id":"TEE/kimi-k3","name":"Kimi K3 TEE","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65535},"cost":{"input":3,"output":15,"cache_read":1.5},"canonical_model_id":"moonshotai/kimi-k3"},"TEE/glm-5.1":{"id":"TEE/glm-5.1","name":"GLM 5.1 TEE","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"input":202752,"output":65535},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"},"TEE/nemotron-3.5-lightning":{"id":"TEE/nemotron-3.5-lightning","name":"Nvidia Nemotron 3.5 Lightning TEE","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.08,"output":0.2,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"TEE/kimi-k2.7-code":{"id":"TEE/kimi-k2.7-code","name":"Kimi K2.7 Code TEE","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"TEE/glm-5.2:thinking":{"id":"TEE/glm-5.2:thinking","name":"GLM 5.2 Thinking TEE","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.7},"canonical_model_id":"zhipuai/glm-5.2"},"TEE/qwen3.5-397b-a17b":{"id":"TEE/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B TEE","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.55,"output":3.5,"cache_read":0.275},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"TEE/gemma-4-26b-a4b-uncensored":{"id":"TEE/gemma-4-26b-a4b-uncensored","name":"Gemma 4 26B A4B Uncensored TEE","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-05-23","last_updated":"2026-05-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":65536},"cost":{"input":0.15,"output":0.7,"cache_read":0.075}},"TEE/kimi-k2.6":{"id":"TEE/kimi-k2.6","name":"Kimi K2.6 TEE","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":1.5,"output":5.25,"cache_read":0.375},"canonical_model_id":"moonshotai/kimi-k2.6"},"TEE/qwen3.6-27b":{"id":"TEE/qwen3.6-27b","name":"Qwen3.6 27B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.32,"output":2.7,"cache_read":0.16},"canonical_model_id":"alibaba/qwen3.6-27b"},"TEE/glm-5.2":{"id":"TEE/glm-5.2","name":"GLM 5.2 TEE","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.7},"canonical_model_id":"zhipuai/glm-5.2"},"TEE/glm-5.3-flash":{"id":"TEE/glm-5.3-flash","name":"GLM 5.3 Flash TEE","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"TEE/gpt-oss-120b":{"id":"TEE/gpt-oss-120b","name":"GPT-OSS 120B TEE","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":2,"output":2,"cache_read":2},"canonical_model_id":"openai/gpt-oss-120b"},"TEE/llama3-3-70b":{"id":"TEE/llama3-3-70b","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":1.75,"output":2.75,"cache_read":1.75},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"TEE/gemma4-31b":{"id":"TEE/gemma4-31b","name":"Gemma 4 31B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.4,"output":1,"cache_read":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"TEE/gemma-4-31b-it":{"id":"TEE/gemma-4-31b-it","name":"Gemma 4 31B IT TEE","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.15,"output":0.46,"cache_read":0.075},"canonical_model_id":"google/gemma-4-31b-it"},"TEE/qwen3.5-27b":{"id":"TEE/qwen3.5-27b","name":"Qwen3.5 27B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.3,"output":2.4,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.5-27b"},"TEE/qwen3.6-35b-a3b":{"id":"TEE/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B TEE","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.2,"output":1.27,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"TEE/deepseek-v3.2":{"id":"TEE/deepseek-v3.2","name":"DeepSeek V3.2 TEE","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"input":164000,"output":65536},"cost":{"input":0.5,"output":1,"cache_read":0.25},"canonical_model_id":"deepseek/deepseek-v3.2"},"TEE/glm-5.1-thinking":{"id":"TEE/glm-5.1-thinking","name":"GLM 5.1 Thinking TEE","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"input":202752,"output":65535},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"}}},"nan":{"id":"nan","env":["NAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.nan.builders/v1","name":"NaN","doc":"https://nan.builders/docs/models","models":{"qwen3.6":{"id":"qwen3.6","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"qwen-image-2.1":{"id":"qwen-image-2.1","type":"image","name":"Qwen-Image-2.1","description":"Unified text-to-image generation and image editing model with native RGBA transparency and reference-image editing","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-14","last_updated":"2026-09-21","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":8192,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen-image-2.1"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm5.3":{"id":"glm5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"glm5.3-flash":{"id":"glm5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemma4":{"id":"gemma4","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-26b-a4b-it"}}},"prime":{"id":"prime","env":["PRIME_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.pinference.ai/api/v1","name":"Prime Inference","doc":"https://docs.primeintellect.ai/inference/overview","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro 0423","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.91,"output":3.83},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash 0423","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.44,"output":1.32},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":1.6,"output":4.8},"canonical_model_id":"xai/grok-4.7"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.119,"output":0.3},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["audio","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["audio","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.35,"output":6.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-3.5-flash"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra Pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.2,"output":1.2},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o-mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.25,"output":2},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.2,"output":1.25},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.05,"output":0.4},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna Pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":1,"output":6},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.75,"output":4.5},"canonical_model_id":"openai/gpt-5.4-mini"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":3.45,"output":17.25},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25},"canonical_model_id":"meta/muse-spark-open"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral Nemo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.165,"output":0.17},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3 2512","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.165,"output":0.66},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2.2,"output":6.6},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":235929},"cost":{"input":0.99,"output":3.2},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.13,"output":0.52},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5-1"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.98,"output":3.08,"cache_read":0.182,"cache_write":0.98},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.54,"output":4.84},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0.15},"canonical_model_id":"zhipuai/glm-5.3-flash"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1.04,"output":2.253},"canonical_model_id":"meta/llama-3.3-70b-instruct"}}},"inferx":{"id":"inferx","env":["INFERX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://model.inferx.net/endpoints/v1","name":"InferX","doc":"https://model.inferx.net/endpoints","models":{"Qwen3-Coder-Next-FP8-no-thinking":{"id":"Qwen3-Coder-Next-FP8-no-thinking","name":"Qwen3-Coder-Next-FP8-no-thinking","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":260000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-next"},"gemma-4-31B-it-fp8":{"id":"gemma-4-31B-it-fp8","name":"Gemma 4 31B IT FP8","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"Qwen3.6-35B-A3B-FP8":{"id":"Qwen3.6-35B-A3B-FP8","name":"Qwen3.6 35B A3B FP8","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Devstral-2-123B-Instruct-2512-int4-AutoRound":{"id":"Devstral-2-123B-Instruct-2512-int4-AutoRound","name":"Devstral-2-123B-Instruct-2512-int4-AutoRound","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"mistral/devstral-2512"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"deepseek-v4-flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":100000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"mimo-v25":{"id":"mimo-v25","name":"mimo-v25","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":100000},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"Qwen3.6-27B-FP8":{"id":"Qwen3.6-27B-FP8","name":"Qwen3.6 27B FP8","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"Agents-A1":{"id":"Agents-A1","name":"Agents-A1","description":"35B MoE agentic model built for long-horizon search, engineering, and scientific reasoning tasks","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"release_date":"2026-06-26","last_updated":"2026-06-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":100000},"cost":{"input":0,"output":0}},"Qwen3-Embedding-8B":{"id":"Qwen3-Embedding-8B","type":"embedding","name":"Qwen3-Embedding-8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"Qwen3-Coder-Next-FP8":{"id":"Qwen3-Coder-Next-FP8","name":"Qwen3 Coder Next FP8","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-next"},"Ornith-1.0-35B-FP8":{"id":"Ornith-1.0-35B-FP8","name":"Ornith-1.0-35B-FP8","description":"Large coding-reasoning model for agentic software tasks and RL search","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-25","last_updated":"2026-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":100000},"cost":{"input":0,"output":0},"canonical_model_id":"deepreinforce/ornith-1.0-35b"},"Qwen3.6-35B-A3B-fp8-no-thinking":{"id":"Qwen3.6-35B-A3B-fp8-no-thinking","name":"Qwen3.6-35B-A3B-fp8-no-thinking","description":"Qwen3.6-35B-A3B-fp8 disable thinking","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"inception":{"id":"inception","env":["INCEPTION_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inceptionlabs.ai/v1/","name":"Inception","doc":"https://docs.inceptionlabs.ai/get-started/models","models":{"mercury-3":{"id":"mercury-3","name":"Mercury 3","description":"Inception's diffusion reasoning model for chat, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01},"canonical_model_id":"inception/mercury-3"},"mercury-2":{"id":"mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"mercury-edit-2":{"id":"mercury-edit-2","name":"Mercury Edit 2","description":"Code editing dLLM for autocomplete (FIM) and next-edit suggestions","family":"mercury","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-30","last_updated":"2026-03-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-edit-2"},"mercury-2.5":{"id":"mercury-2.5","name":"Mercury 2.5","description":"Mercury 2.5 is the fastest reasoning LLM, and the latest diffusion LLM (dLLM) from Inception","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-11-01","release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.2,"output":0.75,"cache_read":0.02},"canonical_model_id":"inception/mercury-2.5"}}},"zhipuai-coding-plan":{"id":"zhipuai-coding-plan","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://open.bigmodel.cn/api/coding/paas/v4","name":"Zhipu AI Coding Plan","doc":"https://docs.bigmodel.cn/cn/coding-plan/overview","models":{"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v"},"glm-5.3-highspeed":{"id":"glm-5.3-highspeed","name":"GLM-5.3 Highspeed","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"}}},"inceptron":{"id":"inceptron","env":["INCEPTRON_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inceptron.io/v1","name":"Inceptron","doc":"https://docs.inceptron.io","models":{"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.11,"output":0.3666,"cache_read":0.027,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6026,"output":3.2,"cache_read":0.169,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.6,"output":2.8,"cache_read":0.12,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.2,"output":0.75,"cache_read":0.075,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.99,"output":1.188,"cache_read":0.16,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"novita-ai":{"id":"novita-ai","env":["NOVITA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.novita.ai/openai","name":"Novita AI","doc":"https://novita.ai/docs/guides/introduction","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-ocr-2":{"id":"deepseek/deepseek-ocr-2","name":"DeepSeek OCR 2","description":"High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0.03},"canonical_model_id":"deepseek/deepseek-ocr-2"},"deepseek/deepseek-r1-0528-qwen3-8b":{"id":"deepseek/deepseek-r1-0528-qwen3-8b","name":"DeepSeek R1 0528 Qwen3 8B","description":"DeepSeek R1 0528 distilled into Qwen3 8B for mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2025-05-29","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.06,"output":0.09},"canonical_model_id":"deepseek/deepseek-r1-0528-qwen3-8b"},"deepseek/deepseek_v3":{"id":"deepseek/deepseek_v3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16000},"cost":{"input":0.89,"output":0.89},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.6,"output":3.2,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.44,"output":1.32,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.7,"output":2.5,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-r1-turbo":{"id":"deepseek/deepseek-r1-turbo","name":"DeepSeek R1 (Turbo)","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16000},"cost":{"input":0.7,"output":2.5},"canonical_model_id":"deepseek/deepseek-r1"},"paddlepaddle/paddleocr-vl":{"id":"paddlepaddle/paddleocr-vl","name":"PaddleOCR VL","description":"Compact vision-language model for multilingual document parsing, OCR, tables, formulas, and charts","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.02,"output":0.02},"canonical_model_id":"paddlepaddle/paddleocr-vl"},"baichuan/baichuan-m2-32b":{"id":"baichuan/baichuan-m2-32b","name":"Baichuan M2 32B","description":"Open medical model trained for clinical reasoning and patient interaction with on, off, and auto thinking modes","family":"baichuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"temperature":true,"release_date":"2025-08-13","last_updated":"2025-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.07,"output":0.07},"canonical_model_id":"baichuan/baichuan-m2-32b"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"inclusionai/ling-3.1-flash"},"inclusionai/ling-3.0-flash-sante":{"id":"inclusionai/ling-3.0-flash-sante","name":"Ling 3.0 Flash Sante","description":"Health and medicine specialization of Ling 3.0 Flash for medical reasoning and evidence-based retrieval","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-sante"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.0-flash-fin":{"id":"inclusionai/ling-3.0-flash-fin","name":"Ling 3.0 Flash Fin","description":"Finance-enhanced model for financial research, multi-step investment workflows, and long-horizon planning and execution","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-27","last_updated":"2026-09-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-fin"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":524288}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"mindai/macaron-v1-tall":{"id":"mindai/macaron-v1-tall","name":"Macaron V1 Tall","description":"Qwen3.6-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2475,"output":1.43,"cache_read":0.044},"canonical_model_id":"mindlab/macaron-v1-tall"},"mindai/macaron-v1-venti":{"id":"mindai/macaron-v1-venti","name":"Macaron V1 Venti","description":"GLM-5.2-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.825,"output":2.475,"cache_read":0.165},"canonical_model_id":"mindlab/macaron-v1-venti"},"apodex/apodex-1.1-mini":{"id":"apodex/apodex-1.1-mini","name":"Apodex 1.1 Mini","description":"Open-weight 35B MoE model for long-horizon research, forecasting, and tool-assisted work","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-24","last_updated":"2026-08-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"apodex/apodex-1.1-mini"},"baidu/ernie-4.5-21B-a3b":{"id":"baidu/ernie-4.5-21B-a3b","name":"ERNIE 4.5 21B A3B","description":"Open ERNIE text MoE with 21B total and 3B active parameters","family":"ernie","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":120000,"output":8000},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"baidu/ernie-4.5-21b-a3b"},"baidu/cobuddy":{"id":"baidu/cobuddy","name":"CoBuddy","description":"Baidu Qianfan model optimized for code generation and tool-using coding agents","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-05-06","last_updated":"2026-05-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":0.28,"output":1.13,"cache_read":0.07},"canonical_model_id":"baidu/cobuddy"},"microsoft/wizardlm-2-8x22b":{"id":"microsoft/wizardlm-2-8x22b","name":"WizardLM 2 8x22B","description":"Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-04-15","last_updated":"2024-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65535,"output":8000},"cost":{"input":0.62,"output":0.62},"canonical_model_id":"microsoft/wizardlm-2-8x22b"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":98304,"output":16384},"cost":{"input":0.119,"output":0.2},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.25},"canonical_model_id":"openai/gpt-oss-120b"},"baai/bge-reranker-v2-m3":{"id":"baai/bge-reranker-v2-m3","type":"reranking","name":"BGE Reranker V2 M3","description":"Lightweight multilingual cross-encoder that scores query-document relevance","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-15","last_updated":"2024-03-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":0},"cost":{"input":0.01,"output":0},"canonical_model_id":"baai/bge-reranker-v2-m3"},"baai/bge-m3":{"id":"baai/bge-m3","type":"embedding","name":"BGE M3","description":"Multilingual embedding model combining dense, sparse, and multi-vector retrieval","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-30","last_updated":"2024-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":0},"cost":{"input":0.01,"output":0},"canonical_model_id":"baai/bge-m3"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2-instruct":{"id":"moonshotai/kimi-k2-instruct","name":"Kimi K2 Instruct","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.8,"output":3.4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 Instruct 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax L2 13B","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":3200},"cost":{"input":0.09,"output":0.09},"canonical_model_id":"gryphe/mythomax-l2-13b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.42,"output":3,"cache_read":0.085},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen-mt-plus":{"id":"qwen/qwen-mt-plus","name":"Qwen MT Plus","description":"Flagship Qwen translation model with terminology control and support for 92 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-07-24","last_updated":"2025-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":0.25,"output":0.75},"canonical_model_id":"alibaba/qwen-mt-plus"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":3.75,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-reranker-8b":{"id":"qwen/qwen3-reranker-8b","type":"reranking","name":"Qwen3 Reranker 8B","description":"Multilingual query-document reranking model supporting more than 100 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0.05,"output":0},"canonical_model_id":"alibaba/qwen3-reranker-8b"},"qwen/qwen3-embedding-8b":{"id":"qwen/qwen3-embedding-8b","type":"embedding","name":"Qwen3 Embedding 8B","description":"Multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0.07,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-omni-30b-a3b-thinking":{"id":"qwen/qwen3-omni-30b-a3b-thinking","name":"Qwen3 Omni 30B A3B Thinking","description":"Multimodal reasoning model accepting text, images, video, and audio with text output","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.25,"output":0.97,"input_audio":2.2},"canonical_model_id":"alibaba/qwen3-omni-30b-a3b-thinking"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"xiaomimimo/mimo-v2.6-flash":{"id":"xiaomimimo/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomimimo/mimo-v2.6-pro":{"id":"xiaomimimo/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomimimo/mimo-v2.5":{"id":"xiaomimimo/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.168,"output":0.336,"cache_read":0.0034},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomimimo/mimo-v2.5-pro":{"id":"xiaomimimo/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.522,"output":1.044,"cache_read":0.0043},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"minimaxai/minimax-m1-80k":{"id":"minimaxai/minimax-m1-80k","name":"MiniMax M1 80K","description":"Open hybrid-attention MoE reasoning model with a million-token context and an 80K thinking budget","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":40000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"minimax/minimax-m1-80k"},"nousresearch/hermes-2-pro-llama-3-8b":{"id":"nousresearch/hermes-2-pro-llama-3-8b","name":"Hermes 2 Pro Llama 3 8B","description":"Llama 3 fine-tune optimized for function calling, structured JSON responses, and dialogue","family":"hermes","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-04-30","last_updated":"2024-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.14,"output":0.14},"canonical_model_id":"nousresearch/hermes-2-pro-llama-3-8b"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":12288,"output":12288},"cost":{"input":0.135,"output":0.4},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama-3.1-8B-Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.02,"output":0.05},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-4-maverick-17b-128e-instruct-fp8":{"id":"meta-llama/llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/llama-4-scout-17b-16e-instruct":{"id":"meta-llama/llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.18,"output":0.59},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.2-1b-instruct":{"id":"meta-llama/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Compact instruction-tuned Llama model for multilingual dialogue and on-device assistants","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":32000},"cost":{"input":0.02,"output":0.02},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"zai-org/glm-5":{"id":"zai-org/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/glm-5.3":{"id":"zai-org/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/glm-5.1":{"id":"zai-org/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/glm-4.5v":{"id":"zai-org/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai-org/glm-4.7-flash":{"id":"zai-org/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/autoglm-phone-9b-multilingual":{"id":"zai-org/autoglm-phone-9b-multilingual","name":"AutoGLM Phone 9B Multilingual","description":"Multilingual visual model for operating smartphone interfaces and mobile agents","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-12-10","last_updated":"2025-12-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.035,"output":0.138},"canonical_model_id":"zhipuai/autoglm-phone-9b-multilingual"},"zai-org/glm-4.7":{"id":"zai-org/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-5.3-flash":{"id":"zai-org/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/glm-4.6":{"id":"zai-org/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/glm-4.6v":{"id":"zai-org/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"zai-org/glm-4.5-air":{"id":"zai-org/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"}}},"ainetcafe":{"id":"ainetcafe","env":["AINETCAFE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://microquickjs.com/v1","name":"ainetcafe","doc":"https://ainetcafe.com/k3/guides/","models":{"Kimi-K3":{"id":"Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":2.1,"output":10.5,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"}}},"digitalocean":{"id":"digitalocean","env":["DIGITALOCEAN_ACCESS_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.do-ai.run/v1","name":"DigitalOcean","doc":"https://docs.digitalocean.com/products/gradient-ai-platform/details/models/","models":{"openai-gpt-6-sol":{"id":"openai-gpt-6-sol","name":"OpenAI GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai-gpt-4o":{"id":"openai-gpt-4o","name":"OpenAI GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"anthropic-claude-opus-4.6":{"id":"anthropic-claude-opus-4.6","name":"Anthropic Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-6"},"openai-gpt-oss-120b":{"id":"openai-gpt-oss-120b","name":"OpenAI GPT-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.1,"output":0.7,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-120b"},"gemma-4-31B-it":{"id":"gemma-4-31B-it","name":"Gemma 4","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.18,"output":0.5,"cache_read":0.036},"canonical_model_id":"google/gemma-4-31b-it"},"gte-large-en-v1.5":{"id":"gte-large-en-v1.5","type":"embedding","name":"GTE Large (v1.5)","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-27","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":1024},"cost":{"input":0.09,"output":0}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"anthropic-claude-opus-4.7":{"id":"anthropic-claude-opus-4.7","name":"Anthropic Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax M2.5 (Public Preview)","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax-m2.5","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08","release_date":"2026-02-12","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.3,"output":1.2,"cache_read":0.06}},"anthropic-claude-opus-4.5":{"id":"anthropic-claude-opus-4.5","name":"Anthropic Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"openai-gpt-4.1":{"id":"openai-gpt-4.1","name":"OpenAI GPT-4.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"Deepseek V4 Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.348},"canonical_model_id":"deepseek/deepseek-v4-pro"},"openai-gpt-6-luna":{"id":"openai-gpt-6-luna","name":"OpenAI GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"nemotron-3.5-super-vl":{"id":"nemotron-3.5-super-vl","name":"NVIDIA Nemotron 3.5 Super VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":0.65,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"deepseek-3.2":{"id":"deepseek-3.2","name":"Deepseek 3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.5,"output":1.6,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v3.2"},"openai-gpt-5.4":{"id":"openai-gpt-5.4","name":"OpenAI GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"qwen3-embedding-0.6b":{"id":"qwen3-embedding-0.6b","type":"embedding","name":"Qwen3 Embedding 0.6B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-03","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":1024},"status":"beta","cost":{"input":0.04,"output":0}},"openai-gpt-5":{"id":"openai-gpt-5","name":"OpenAI GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"qwen3-tts-voicedesign":{"id":"qwen3-tts-voicedesign","type":"speech","name":"Qwen3 TTS VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["audio"]},"open_weights":true,"limit":{"context":32768,"output":1},"provider":{"npm":"@ai-sdk/openai"}},"anthropic-claude-opus-4.8":{"id":"anthropic-claude-opus-4.8","name":"Anthropic Claude Opus 4.8","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"openai-gpt-5-nano":{"id":"openai-gpt-5-nano","name":"OpenAI GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai-gpt-5-mini":{"id":"openai-gpt-5-mini","name":"OpenAI GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai-gpt-5.5":{"id":"openai-gpt-5.5","name":"OpenAI GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai-gpt-5.4-nano":{"id":"openai-gpt-5.4-nano","name":"OpenAI GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai-o1":{"id":"openai-o1","name":"OpenAI o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai-gpt-5.6-terra":{"id":"openai-gpt-5.6-terra","name":"OpenAI GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"wan2-2-t2v-a14b":{"id":"wan2-2-t2v-a14b","type":"video","name":"Wan2.2-T2V-A14B","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-07-28","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["video"]},"open_weights":true,"limit":{"context":100,"output":1}},"anthropic-claude-sonnet-5.5":{"id":"anthropic-claude-sonnet-5.5","name":"Anthropic Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"nemotron-3-ultra-550b":{"id":"nemotron-3-ultra-550b","name":"Nemotron 3 Ultra","description":"Flagship Nemotron model for high-throughput reasoning and complex agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.9,"output":1.7,"cache_read":0.18},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"arcee-trinity-large-thinking":{"id":"arcee-trinity-large-thinking","name":"Arcee Trinity Large Thinking (Public Preview)","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.25,"output":0.9,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"anthropic-claude-haiku-4.5":{"id":"anthropic-claude-haiku-4.5","name":"Anthropic Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic-claude-fable-5.1":{"id":"anthropic-claude-fable-5.1","name":"Anthropic Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"anthropic-claude-5.5-haiku":{"id":"anthropic-claude-5.5-haiku","name":"Anthropic Claude 5.5 Haiku","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"multi-qa-mpnet-base-dot-v1":{"id":"multi-qa-mpnet-base-dot-v1","type":"embedding","name":"Multi-QA-mpnet-base-dot-v1","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2021-08-23","last_updated":"2021-08-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":768},"cost":{"input":0.009,"output":0},"canonical_model_id":"sentence-transformers/multi-qa-mpnet-base-dot-v1"},"openai-gpt-image-2":{"id":"openai-gpt-image-2","type":"image","name":"OpenAI GPT Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image","text"]},"open_weights":false,"limit":{"context":0,"output":16384},"cost":{"input":8,"output":30},"canonical_model_id":"openai/gpt-image-2"},"openai-gpt-5.2":{"id":"openai-gpt-5.2","name":"OpenAI GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"all-mini-lm-l6-v2":{"id":"all-mini-lm-l6-v2","type":"embedding","name":"All-MiniLM-L6-v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2021-08-30","last_updated":"2021-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256,"output":384},"cost":{"input":0.009,"output":0},"canonical_model_id":"sentence-transformers/all-minilm-l6-v2"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen 3.5 397B A17B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.55,"output":3.5,"cache_read":0.111},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"e5-large-v2":{"id":"e5-large-v2","type":"embedding","name":"E5 Large v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-05","last_updated":"2023-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"cost":{"input":0.02,"output":0},"canonical_model_id":"intfloat/e5-large-v2"},"mistral-3-14B":{"id":"mistral-3-14B","name":"Ministral 3 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.2,"output":0.2,"cache_read":0.04},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8-Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-max"},"bge-reranker-v2-m3":{"id":"bge-reranker-v2-m3","type":"reranking","name":"BGE Reranker v2 M3","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-15","last_updated":"2024-03-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":1},"cost":{"input":0.01,"output":0},"canonical_model_id":"baai/bge-reranker-v2-m3"},"openai-gpt-4o-mini":{"id":"openai-gpt-4o-mini","name":"OpenAI GPT-4o mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.6"},"nemotron-3-nano-omni":{"id":"nemotron-3-nano-omni","name":"Nemotron 3 Nano Omni","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":0.9,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"anthropic-claude-4.6-sonnet":{"id":"anthropic-claude-4.6-sonnet","name":"Anthropic Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"openai-gpt-5.6-luna":{"id":"openai-gpt-5.6-luna","name":"OpenAI GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.8,"output":3,"cache_read":0.16},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"openai-gpt-oss-20b":{"id":"openai-gpt-oss-20b","name":"OpenAI GPT-oss-20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.45,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-20b"},"openai-gpt-image-1":{"id":"openai-gpt-image-1","type":"image","name":"GPT Image 1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":40,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.2"},"openai-gpt-6-1-sol":{"id":"openai-gpt-6-1-sol","name":"OpenAI GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"anthropic-claude-opus-5":{"id":"anthropic-claude-opus-5","name":"Anthropic Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"openai-gpt-6-astra":{"id":"openai-gpt-6-astra","name":"OpenAI GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai-o3-mini":{"id":"openai-o3-mini","name":"OpenAI o3 mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"llama-4-maverick":{"id":"llama-4-maverick","name":"Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.25,"output":0.87,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"anthropic-claude-opus-5.5":{"id":"anthropic-claude-opus-5.5","name":"Anthropic Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.4,"cache_write":10,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.4,"cache_write":10}},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic-claude-fable-5":{"id":"anthropic-claude-fable-5","name":"Anthropic Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"openai-gpt-5.3-codex":{"id":"openai-gpt-5.3-codex","name":"OpenAI GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"stable-diffusion-3.5-large":{"id":"stable-diffusion-3.5-large","type":"image","name":"Stable Diffusion 3.5 Large","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"stable-diffusion","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-22","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":256,"output":1}},"reflection-beam":{"id":"reflection-beam","name":"Beam","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.4,"output":1.2,"cache_read":0.04}},"openai-gpt-image-1.5":{"id":"openai-gpt-image-1.5","type":"image","name":"OpenAI GPT Image 1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":16384},"cost":{"input":5,"output":10,"cache_read":1},"canonical_model_id":"openai/gpt-image-1.5"},"bge-m3":{"id":"bge-m3","type":"embedding","name":"BGE M3","description":"Multilingual embedding model combining dense, sparse, and multi-vector retrieval","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-30","last_updated":"2024-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":1024},"cost":{"input":0.02,"output":0},"canonical_model_id":"baai/bge-m3"},"anthropic-claude-4.5-sonnet":{"id":"anthropic-claude-4.5-sonnet","name":"Anthropic Claude 4.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"nemotron-nano-12b-v2-vl":{"id":"nemotron-nano-12b-v2-vl","name":"Nemotron-nano 12b v2-vl","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.2,"output":0.6,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"openai-o3":{"id":"openai-o3","name":"OpenAI o3","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai-gpt-5.2-pro":{"id":"openai-gpt-5.2-pro","name":"OpenAI GPT-5.2 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"deepseek-4-flash":{"id":"deepseek-4-flash","name":"Deepseek V4 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"openai-gpt-5.6-sol":{"id":"openai-gpt-5.6-sol","name":"OpenAI GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai-gpt-5.4-pro":{"id":"openai-gpt-5.4-pro","name":"OpenAI GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai-gpt-5.4-mini":{"id":"openai-gpt-5.4-mini","name":"OpenAI GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"anthropic-claude-5-sonnet":{"id":"anthropic-claude-5-sonnet","name":"Anthropic Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"fal-ai/fast-sdxl":{"id":"fal-ai/fast-sdxl","type":"image","name":"Fast SDXL","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"stable-diffusion","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-07-26","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0}},"fal-ai/elevenlabs/tts/multilingual-v2":{"id":"fal-ai/elevenlabs/tts/multilingual-v2","type":"speech","name":"ElevenLabs Multilingual TTS v2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-08-22","last_updated":"2023-08-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"canonical_model_id":"elevenlabs/eleven_multilingual_v2"},"fal-ai/flux/schnell":{"id":"fal-ai/flux/schnell","type":"image","name":"FLUX.1 [schnell]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-01","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0}},"fal-ai/stable-audio-25/text-to-audio":{"id":"fal-ai/stable-audio-25/text-to-audio","type":"speech","name":"Stable Audio 2.5 (Text-to-Audio)","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-08","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0}}}},"model-oracle-ai":{"id":"model-oracle-ai","env":["MODEL_ORACLE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.modeloracle.com/api/v1","name":"Model Oracle AI","doc":"https://modeloracle.com/setup/","models":{"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-pro"},"auto":{"id":"auto","name":"Auto","description":"Model Oracle AI decision engine that selects and routes among configured coding-agent models","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-29","last_updated":"2026-07-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-5.4"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"canonical_model_id":"anthropic/claude-fable-5"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1-mini"},"claude-haiku-4.5":{"id":"claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"canonical_model_id":"anthropic/claude-haiku-4-5"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.4-nano"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"canonical_model_id":"anthropic/claude-sonnet-5"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o4-mini"},"claude-opus-4.8":{"id":"claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-5.5"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.4-mini"}}},"ovhcloud":{"id":"ovhcloud","env":["OVHCLOUD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://oai.endpoints.kepler.ai.cloud.ovh.net/v1","name":"OVHcloud AI Endpoints","doc":"https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog//","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8-27B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.47,"output":3.19},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen2.5-vl-72b-instruct":{"id":"qwen2.5-vl-72b-instruct","name":"Qwen2.5-VL-72B-Instruct","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":1.01,"output":1.01},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"qwen3guard-gen-8b":{"id":"qwen3guard-gen-8b","name":"Qwen3Guard-Gen-8B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-22","last_updated":"2026-01-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0,"output":0}},"mistral-nemo-instruct-2407":{"id":"mistral-nemo-instruct-2407","name":"Mistral-Nemo-Instruct-2407","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.14,"output":0.14},"canonical_model_id":"mistral/mistral-nemo"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"gpt-oss-20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5-9B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.12,"output":0.18},"canonical_model_id":"alibaba/qwen3.5-9b"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"Mistral-Small-3.2-24B-Instruct-2506","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.31},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5-397B-A17B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.71,"output":4.25},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6-27B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.47,"output":3.19},"canonical_model_id":"alibaba/qwen3.6-27b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"gpt-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.09,"output":0.47},"canonical_model_id":"openai/gpt-oss-120b"},"qwen3guard-gen-0.6b":{"id":"qwen3guard-gen-0.6b","name":"Qwen3Guard-Gen-0.6B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-22","last_updated":"2026-01-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0,"output":0}},"meta-llama-3_3-70b-instruct":{"id":"meta-llama-3_3-70b-instruct","name":"Meta-Llama-3_3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.74,"output":0.74},"canonical_model_id":"meta/llama-3.3-70b-instruct"}}},"sarvam":{"id":"sarvam","env":["SARVAM_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.sarvam.ai/v1","name":"Sarvam AI","doc":"https://docs.sarvam.ai/api-reference-docs/getting-started/models","models":{"sarvam-105b":{"id":"sarvam-105b","name":"Sarvam-105B","description":"Flagship Indian-language reasoning model for enterprise multilingual applications","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.15790269,"output":0.63161075,"cache_read":0.05263423},"canonical_model_id":"sarvam/sarvam-105b"},"sarvam-30b":{"id":"sarvam-30b","name":"Sarvam-30B","description":"Efficient Indian-language reasoning model for chat, coding, and multilingual work","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"status":"deprecated","canonical_model_id":"sarvam/sarvam-30b"}}},"302ai":{"id":"302ai","env":["302AI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.302.ai/v1","name":"302.AI","doc":"https://doc.302.ai","models":{"glm-5":{"id":"glm-5","name":"glm-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.6,"tiers":[{"input":0.9,"output":3.2,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"claude-haiku-4-5-20251001","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"claude-opus-4-1-20250805-thinking":{"id":"claude-opus-4-1-20250805-thinking","name":"claude-opus-4-1-20250805-thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"glm-5-turbo":{"id":"glm-5-turbo","name":"glm-5-turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.72,"output":3.2,"tiers":[{"input":1.1,"output":3.8,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5-turbo"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-v3.2-thinking":{"id":"deepseek-v3.2-thinking","name":"DeepSeek-V3.2-Thinking","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.29,"output":0.43},"canonical_model_id":"deepseek/deepseek-v3.2"},"gpt-5.6-terra-pro":{"id":"gpt-5.6-terra-pro","name":"gpt-5.6-terra-pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-terra"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"gemini-2.5-flash-lite-preview-09-2025":{"id":"gemini-2.5-flash-lite-preview-09-2025","name":"gemini-2.5-flash-lite-preview-09-2025","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.8,"tiers":[{"input":1.2,"output":7.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":7.2}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen3-235B-A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.29,"output":2.86},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5},"canonical_model_id":"xai/grok-4.3"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"claude-opus-4-5-20251101","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.29,"output":0.86,"tiers":[{"input":0.58,"output":1.72,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5v"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"gemini-2.5-pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":10,"tiers":[{"input":2.5,"output":15,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15}},"canonical_model_id":"google/gemini-2.5-pro"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"claude-haiku-4-5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5},"canonical_model_id":"anthropic/claude-haiku-4-5"},"gpt-5.2-chat-latest":{"id":"gpt-5.2-chat-latest","name":"gpt-5.2-chat-latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"mistral-large-2512":{"id":"mistral-large-2512","name":"mistral-large-2512","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"cost":{"input":1.1,"output":3.3},"canonical_model_id":"mistral/mistral-large-2512"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.8,"output":5.3},"canonical_model_id":"alibaba/qwen3.7-max"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.627,"output":3.3},"canonical_model_id":"moonshotai/kimi-k2.5"},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"ministral-14b-2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.33,"output":0.33},"canonical_model_id":"mistral/ministral-14b-2512"},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"gemini-3-pro-image-preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":64000},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image-preview"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02}},"canonical_model_id":"openai/gpt-6-luna"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"gpt-4.1-nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"openai/gpt-4.1-nano"},"qwen3-max-2025-09-23":{"id":"qwen3-max-2025-09-23","name":"qwen3-max-2025-09-23","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":258048,"output":65536},"cost":{"input":0.86,"output":3.43,"tiers":[{"input":1.43,"output":5.72,"tier":{"type":"context","size":32000}},{"input":2.15,"output":8.58,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"qwen3-235b-a22b-instruct-2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":65536},"cost":{"input":0.29,"output":1.143},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"gpt-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":0,"tiers":[{"input":5,"output":22.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5}},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.3-chat-latest":{"id":"gpt-5.3-chat-latest","name":"GPT-5.3 Chat (latest)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.06,"output":0.46,"tiers":[{"input":0.23,"output":1.83,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"doubao-seed-1-6-vision-250815":{"id":"doubao-seed-1-6-vision-250815","name":"doubao-seed-1-6-vision-250815","description":"ByteDance Seed multimodal model for image understanding, visual reasoning, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-08-15","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.114,"output":1.143},"canonical_model_id":"bytedance-seed/seed-1-6-vision"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.33,"output":1.32},"canonical_model_id":"minimax/MiniMax-M2"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"claude-sonnet-4-5-20250929","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"qwen3.7-max-2026-06-08":{"id":"qwen3.7-max-2026-06-08","name":"qwen3.7-max-2026-06-08","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.8,"output":5.3},"canonical_model_id":"alibaba/qwen3.7-max"},"doubao-seed-1-8-251215":{"id":"doubao-seed-1-8-251215","name":"doubao-seed-1-8-251215","description":"ByteDance Seed model for multimodal reasoning, long-context analysis, and agent workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":224000,"output":64000},"cost":{"input":0.1143,"output":0.286,"tiers":[{"input":0.1715,"output":2.286,"tier":{"type":"context","size":32000}},{"input":0.343,"output":3.43,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-8"},"glm-4.7":{"id":"glm-4.7","name":"glm-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.286,"output":1.142,"tiers":[{"input":0.572,"output":2.29,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":10,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"claude-opus-4-7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-7"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.286,"output":1.143,"tiers":[{"input":0.5714,"output":2.29,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.72,"output":2.88,"cache_read":0.144,"cache_write":0.9,"tiers":[{"input":1.44,"output":5.76,"cache_read":0.144,"cache_write":0.9,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":1.44,"output":5.76,"cache_read":0.144,"cache_write":0.9}},"canonical_model_id":"minimax/MiniMax-M3"},"gpt-5.2":{"id":"gpt-5.2","name":"gpt-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"gpt-4.1-mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"openai/gpt-4.1-mini"},"gpt-5-mini":{"id":"gpt-5-mini","name":"gpt-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1},"canonical_model_id":"openai/gpt-6-astra"},"gemini-3.5-flash-thinking":{"id":"gemini-3.5-flash-thinking","name":"gemini-3.5-flash-thinking","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.1-chat-latest":{"id":"gpt-5.1-chat-latest","name":"gpt-5.1-chat-latest","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"claude-opus-4-1-20250805","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gpt-5-pro":{"id":"gpt-5-pro","name":"gpt-5-pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-pro"},"grok-4-fast-reasoning":{"id":"grok-4-fast-reasoning","name":"grok-4-fast-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4-fast-reasoning"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.16,"output":6.36,"cache_read":0.168,"cache_write":2.58},"canonical_model_id":"alibaba/qwen3.8-max"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"tiers":[{"input":4,"output":18,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":0.5},"canonical_model_id":"xai/grok-4.7"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"gpt-5.1":{"id":"gpt-5.1","name":"gpt-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1"},"claude-sonnet-4-5-20250929-thinking":{"id":"claude-sonnet-4-5-20250929-thinking","name":"claude-sonnet-4-5-20250929-thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gemini-2.5-flash-preview-09-2025":{"id":"gemini-2.5-flash-preview-09-2025","name":"gemini-2.5-flash-preview-09-2025","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5}},"gpt-5-thinking":{"id":"gpt-5-thinking","name":"gpt-5-thinking","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5"},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.12,"output":0.69,"tiers":[{"input":0.29,"output":1.72,"tier":{"type":"context","size":128000}},{"input":0.58,"output":3.44,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.5-plus"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6-thinking":{"id":"claude-sonnet-4-6-thinking","name":"claude-sonnet-4-6-thinking","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"gemini-3-flash-preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3-flash-preview"},"gpt-5.6-sol-pro":{"id":"gpt-5.6-sol-pro","name":"gpt-5.6-sol-pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.6-sol"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"gpt-5.4-nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"grok-4.1":{"id":"grok-4.1","name":"grok-4.1","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":2,"output":10}},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-4o":{"id":"gpt-4o","name":"gpt-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-4o"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.075,"output":0.25,"cache_read":0.015},"canonical_model_id":"zhipuai/glm-5.3-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.228,"output":0.92,"cache_read":0.023,"cache_write":0.286,"tiers":[{"input":0.68,"output":2.74,"cache_read":0.068,"cache_write":0.86,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.68,"output":2.74,"cache_read":0.068,"cache_write":0.86}},"canonical_model_id":"alibaba/qwen3.7-plus"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"gemini-2.5-flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash"},"gemini-2.5-flash-nothink":{"id":"gemini-2.5-flash-nothink","name":"gemini-2.5-flash-nothink","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash"},"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.72,"output":3.2,"tiers":[{"input":1.1,"output":3.8,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"glm-4.6":{"id":"glm-4.6","name":"glm-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.286,"output":1.142,"tiers":[{"input":0.572,"output":2.29,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.145,"output":0.43,"tiers":[{"input":0.29,"output":0.86,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3-30B-A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.11,"output":1.08},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"grok-4-1-fast-reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"gemini-2.0-flash-lite":{"id":"gemini-2.0-flash-lite","name":"gemini-2.0-flash-lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-11","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":8192},"cost":{"input":0.075,"output":0.3},"canonical_model_id":"google/gemini-2.0-flash-lite"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"kimi-k2-thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.575,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"MiniMax-M1":{"id":"MiniMax-M1","name":"MiniMax-M1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-16","last_updated":"2025-06-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.132,"output":1.254}},"doubao-seed-1-6-thinking-250715":{"id":"doubao-seed-1-6-thinking-250715","name":"doubao-seed-1-6-thinking-250715","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16000},"cost":{"input":0.121,"output":1.21}},"gpt-5":{"id":"gpt-5","name":"gpt-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5"},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"qwen3-coder-480b-a35b-instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.86,"output":3.43,"tiers":[{"input":1.29,"output":5.15,"tier":{"type":"context","size":32000}},{"input":2.15,"output":8.58,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8},"canonical_model_id":"openai/o3"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.5"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"claude-sonnet-4-6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"grok-4.20-beta-0309-reasoning":{"id":"grok-4.20-beta-0309-reasoning","name":"grok-4.20-beta-0309-reasoning","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"grok-4-fast-non-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"gemini-3.1-flash-image-preview":{"id":"gemini-3.1-flash-image-preview","type":"image","name":"gemini-3.1-flash-image-preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":60},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"gpt-5.6-luna-pro":{"id":"gpt-5.6-luna-pro","name":"gpt-5.6-luna-pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-3.5-flash"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.28,"output":1.7},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4}},"canonical_model_id":"openai/gpt-6-sol"},"claude-opus-4-7-thinking":{"id":"claude-opus-4-7-thinking","name":"claude-opus-4-7-thinking","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-7"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"deepseek-v3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.29,"output":0.43},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-flash":{"id":"deepseek-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"gpt-4.1":{"id":"gpt-4.1","name":"gpt-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8},"canonical_model_id":"openai/gpt-4.1"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"gemini-2.5-flash-image","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30},"canonical_model_id":"google/gemini-2.5-flash-image"},"kimi-k2-0905-preview":{"id":"kimi-k2-0905-preview","name":"kimi-k2-0905-preview","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.62854,"output":2.5146},"canonical_model_id":"moonshotai/kimi-k2-0905"},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.19,"output":1.13,"cache_read":0.0171,"cache_write":0.214,"tiers":[{"input":0.76,"output":4.52,"cache_read":0.0686,"cache_write":0.857,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.76,"output":4.52,"cache_read":0.0686,"cache_write":0.857}},"canonical_model_id":"alibaba/qwen3.6-flash"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"gemini-3-pro-preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":12,"tiers":[{"input":4,"output":18,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18}},"canonical_model_id":"google/gemini-3-pro-preview"},"claude-opus-5-thinking":{"id":"claude-opus-5-thinking","name":"claude-opus-5-thinking","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"gpt-5.4-mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"scnet-token-plan":{"id":"scnet-token-plan","env":["SCNET_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.scnet.cn/api/llm/v1","name":"SCNet Token Plan","doc":"https://www.scnet.cn/ac/openapi/doc/2.0/moduleapi/plans/token-plan.html","models":{"DeepSeek-V4-Flash-0731":{"id":"DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"GLM-5.3":{"id":"GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3"},"Qwen3.8-Max":{"id":"Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-max"},"Kimi-K2.5":{"id":"Kimi-K2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"Kimi-K3":{"id":"Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k3"},"DeepSeek-V4-Pro-0813":{"id":"DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.1"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.7"},"DeepSeek-V4.1-Flash":{"id":"DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3"},"GLM-5":{"id":"GLM-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5"},"Qwen3.8-Flash":{"id":"Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"Kimi-K2.7-Code":{"id":"Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.5"},"GLM-5.3-Flash":{"id":"GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"routing-run":{"id":"routing-run","env":["ROUTING_RUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.routing.run/v1","name":"routing.run","doc":"https://docs.routing.run/api-reference/models","models":{"nemotron-3-ultra":{"id":"nemotron-3-ultra","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32000},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":1.5,"output":9},"canonical_model_id":"openai/gpt-5.6-terra"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.348,"output":0.696},"canonical_model_id":"deepseek/deepseek-v4-pro"},"kimi-k2.6-nitro":{"id":"kimi-k2.6-nitro","name":"Kimi K2.6 Nitro","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":0.7,"output":4.2},"canonical_model_id":"openai/gpt-5.6-luna"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32000},"cost":{"input":0.16,"output":0.48},"canonical_model_id":"alibaba/qwen3.5-9b"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.112,"output":0.224},"canonical_model_id":"deepseek/deepseek-v4-flash"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.6"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.8,"output":2.4},"canonical_model_id":"zhipuai/glm-5.2"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"kimi-k2.7-code-nitro":{"id":"kimi-k2.7-code-nitro","name":"Kimi K2.7 Code Nitro","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-5.2-nitro":{"id":"glm-5.2-nitro","name":"GLM 5.2 Nitro","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.8,"output":2.4},"canonical_model_id":"zhipuai/glm-5.2"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.6-sol"}}},"vultr":{"id":"vultr","env":["VULTR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.vultrinference.com/v1","name":"Vultr","doc":"https://api.vultrinference.com/","models":{"laguna-s-2.1":{"id":"laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.09,"output":0.18},"canonical_model_id":"poolside/laguna-s-2.1"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":1},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.75,"output":3},"canonical_model_id":"zhipuai/glm-5.3"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.25,"output":1},"canonical_model_id":"meta/muse-glimmer-30b"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"deprecated","cost":{"input":0.1,"output":0.25},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"nemotron-3-nano-omni-30b-a3b-reasoning":{"id":"nemotron-3-nano-omni-30b-a3b-reasoning","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.25},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.2,"output":0.9},"canonical_model_id":"minimax/MiniMax-M3"},"nemotron-3.5-content-safety":{"id":"nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.75,"output":3},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.8-flash-next":{"id":"qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.35},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5.x-menthol":{"id":"glm-5.x-menthol","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.4,"output":1.75},"canonical_model_id":"zhipuai/glm-5"},"mimo-v2.6-pro-rl":{"id":"mimo-v2.6-pro-rl","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":0.8},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.6-flash-rl":{"id":"mimo-v2.6-flash-rl","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.25},"canonical_model_id":"xiaomi/mimo-v2.6-flash"}}},"morph":{"id":"morph","env":["MORPH_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.morphllm.com/v1","name":"Morph","doc":"https://docs.morphllm.com/api-reference/introduction","models":{"morph-v3-fast":{"id":"morph-v3-fast","name":"Morph v3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":0.8,"output":1.2}},"auto":{"id":"auto","name":"Auto","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.85,"output":1.55}},"morph-v3-large":{"id":"morph-v3-large","name":"Morph v3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.9,"output":1.9}}}},"evroc":{"id":"evroc","env":["EVROC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://models.think.evroc.com/v1","name":"evroc","doc":"https://docs.evroc.com/products/think/overview.html","models":{"evroc/roc":{"id":"evroc/roc","name":"roc","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-06-06","last_updated":"2026-06-06","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":2.875,"output":11.516}},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.345,"output":1.38},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8-27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.87,"output":3.5},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3-Embedding-8B":{"id":"Qwen/Qwen3-Embedding-8B","type":"embedding","name":"Qwen3 Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":4096},"cost":{"input":0.115,"output":0.115},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"Qwen/Qwen3-Reranker-4B":{"id":"Qwen/Qwen3-Reranker-4B","type":"reranking","name":"Qwen3 Reranker 4B","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.0575,"output":0}},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.144,"output":0.575},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"intfloat/multilingual-e5-large-instruct":{"id":"intfloat/multilingual-e5-large-instruct","type":"embedding","name":"E5 Multi-Lingual Large Embeddings 0.6B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"cost":{"input":0.114,"output":0.114}},"openai/whisper-large-v3":{"id":"openai/whisper-large-v3","type":"transcription","name":"Whisper 3 Large","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.0023,"output":0.0023,"output_audio":2.3},"canonical_model_id":"openai/whisper-large-v3"},"openai/whisper-large-v3-turbo":{"id":"openai/whisper-large-v3-turbo","type":"transcription","name":"Whisper Large v3 Turbo","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.0023,"output":0.0023,"output_audio":2.3},"canonical_model_id":"openai/whisper-large-v3-turbo"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.23,"output":0.92},"canonical_model_id":"openai/gpt-oss-120b"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.4375,"output":5.75},"canonical_model_id":"moonshotai/kimi-k2.6"},"mistralai/Mistral-Medium-3.5-128B":{"id":"mistralai/Mistral-Medium-3.5-128B","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.725,"output":6.9},"canonical_model_id":"mistral/mistral-medium-2604"},"mistralai/Voxtral-Small-24B-2507":{"id":"mistralai/Voxtral-Small-24B-2507","type":"transcription","name":"Voxtral Small 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"release_date":"2025-03-01","last_updated":"2025-03-01","modalities":{"input":["audio","text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"provider":{"npm":"@ai-sdk/openai"}},"nvidia/Llama-3.3-70B-Instruct-FP8":{"id":"nvidia/Llama-3.3-70B-Instruct-FP8","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":1.15,"output":1.15},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"KBLab/kb-whisper-large":{"id":"KBLab/kb-whisper-large","type":"transcription","name":"KB Whisper","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"provider":{"npm":"@ai-sdk/openai"}},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":131072},"cost":{"input":1.4375,"output":5.75},"canonical_model_id":"zhipuai/glm-5.2"}}},"amazon-bedrock":{"id":"amazon-bedrock","env":["AWS_ACCESS_KEY_ID","AWS_SECRET_ACCESS_KEY","AWS_REGION","AWS_BEARER_TOKEN_BEDROCK"],"npm":"@ai-sdk/amazon-bedrock","name":"Amazon Bedrock","doc":"https://docs.aws.amazon.com/bedrock/latest/userguide/models-supported.html","models":{"us.anthropic.claude-sonnet-5":{"id":"us.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (US)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"eu.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"eu.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"qwen.qwen3-coder-next":{"id":"qwen.qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"us.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"us.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (US)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"us.meta.llama3-1-8b-instruct-v1:0":{"id":"us.meta.llama3-1-8b-instruct-v1:0","name":"Llama 3.1 8B Instruct (US)","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta.llama4-maverick-17b-instruct-v1:0":{"id":"meta.llama4-maverick-17b-instruct-v1:0","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":8192},"status":"deprecated","cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"us.anthropic.claude-sonnet-5-5":{"id":"us.anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5 (US)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"amazon.nova-lite-v1:0":{"id":"amazon.nova-lite-v1:0","name":"Nova Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015,"cache_write":0.06},"canonical_model_id":"amazon/nova-lite"},"au.anthropic.claude-haiku-5-5":{"id":"au.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (AU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"global.anthropic.claude-sonnet-4-6":{"id":"global.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (Global)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"amazon.nova-pro-v1:0":{"id":"amazon.nova-pro-v1:0","name":"Nova Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2,"cache_write":0.8},"canonical_model_id":"amazon/nova-pro"},"eu.anthropic.claude-sonnet-5-5":{"id":"eu.anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5 (EU)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"openai.gpt-5.6-sol":{"id":"openai.gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5,"tiers":[{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11}},"canonical_model_id":"openai/gpt-5.6-sol"},"us.amazon.nova-pro-v1:0":{"id":"us.amazon.nova-pro-v1:0","name":"Nova Pro (US)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2,"cache_write":0.8},"canonical_model_id":"amazon/nova-pro"},"writer.palmyra-x5-v1:0":{"id":"writer.palmyra-x5-v1:0","name":"Palmyra X5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"input":1040000,"output":8192},"status":"deprecated","cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"global.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"global.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (Global)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"deepseek.r1-v1:0":{"id":"deepseek.r1-v1:0","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"us.zai.glm-5.3":{"id":"us.zai.glm-5.3","name":"GLM-5.3 (US)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.848,"output":5.808,"cache_read":0.3432,"cache_write":2.31},"canonical_model_id":"zhipuai/glm-5.3"},"us.amazon.nova-micro-v1:0":{"id":"us.amazon.nova-micro-v1:0","name":"Nova Micro (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875,"cache_write":0.035},"canonical_model_id":"amazon/nova-micro"},"us.meta.llama4-maverick-17b-instruct-v1:0":{"id":"us.meta.llama4-maverick-17b-instruct-v1:0","name":"Llama 4 Maverick 17B Instruct (US)","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":8192},"status":"deprecated","cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"anthropic.claude-sonnet-4-6":{"id":"anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"nvidia.nemotron-nano-3-30b":{"id":"nvidia.nemotron-nano-3-30b","name":"NVIDIA Nemotron Nano 3 30B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"openai.gpt-6-sol":{"id":"openai.gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6-sol"},"anthropic.claude-fable-5":{"id":"anthropic.claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"nvidia.nemotron-nano-9b-v2":{"id":"nvidia.nemotron-nano-9b-v2","name":"NVIDIA Nemotron Nano 9B v2","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-08-18","last_updated":"2025-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.06,"output":0.23},"canonical_model_id":"nvidia/nemotron-nano-9b-v2"},"meta.llama3-1-70b-instruct-v1:0":{"id":"meta.llama3-1-70b-instruct-v1:0","name":"Llama 3.1 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"qwen.qwen3-next-80b-a3b":{"id":"qwen.qwen3-next-80b-a3b","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-11","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262000},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"global.anthropic.claude-fable-5":{"id":"global.anthropic.claude-fable-5","name":"Claude Fable 5 (Global)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"us.anthropic.claude-haiku-5-5":{"id":"us.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (US)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"in.anthropic.claude-sonnet-5":{"id":"in.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (India)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"meta.llama3-3-70b-instruct-v1:0":{"id":"meta.llama3-3-70b-instruct-v1:0","name":"Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"mistral.ministral-3-14b-instruct":{"id":"mistral.ministral-3-14b-instruct","name":"Ministral 14B 3.0","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"qwen.qwen3-235b-a22b-2507-v1:0":{"id":"qwen.qwen3-235b-a22b-2507-v1:0","name":"Qwen3 235B-A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.22,"output":0.88},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"deepseek.v3-v1:0":{"id":"deepseek.v3-v1:0","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":81920},"status":"deprecated","cost":{"input":0.58,"output":1.68},"canonical_model_id":"deepseek/deepseek-v3.1"},"apac.amazon.nova-lite-v1:0":{"id":"apac.amazon.nova-lite-v1:0","name":"Nova Lite (APAC)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.063,"output":0.252,"cache_read":0.01575,"cache_write":0.063},"canonical_model_id":"amazon/nova-lite"},"au.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"au.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (AU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"jp.anthropic.claude-haiku-5-5":{"id":"jp.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (JP)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"global.openai.gpt-5.6-luna":{"id":"global.openai.gpt-5.6-luna","name":"GPT-5.6 Luna (Global)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"eu.mistral.pixtral-large-2502-v1:0":{"id":"eu.mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02) (EU)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"minimax.minimax-m2":{"id":"minimax.minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204608,"output":128000},"status":"deprecated","cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"eu.amazon.nova-2-lite-v1:0":{"id":"eu.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (EU)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.374,"output":3.157,"cache_read":0.0935,"cache_write":0.374},"canonical_model_id":"amazon/nova-2-lite"},"eu.anthropic.claude-opus-4-8":{"id":"eu.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (EU)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"us.openai.gpt-5.6-terra":{"id":"us.openai.gpt-5.6-terra","name":"GPT-5.6 Terra (US)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"us.xai.grok-4.6":{"id":"us.xai.grok-4.6","name":"Grok 4.6 (US)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.2,"output":6.6,"cache_read":0.55},"canonical_model_id":"xai/grok-4.6"},"anthropic.claude-sonnet-5":{"id":"anthropic.claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"ca.amazon.nova-lite-v1:0":{"id":"ca.amazon.nova-lite-v1:0","name":"Nova Lite (CA)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.064,"output":0.256,"cache_read":0.016,"cache_write":0.064},"canonical_model_id":"amazon/nova-lite"},"google.gemma-3-12b-it":{"id":"google.gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"xai.grok-4.3":{"id":"xai.grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-06-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"meta.llama3-1-8b-instruct-v1:0":{"id":"meta.llama3-1-8b-instruct-v1:0","name":"Llama 3.1 8B Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"moonshotai.kimi-k2.5":{"id":"moonshotai.kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-02-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"eu.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"eu.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"anthropic.claude-opus-5":{"id":"anthropic.claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"eu.amazon.nova-micro-v1:0":{"id":"eu.amazon.nova-micro-v1:0","name":"Nova Micro (EU)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.04,"output":0.16,"cache_read":0.01,"cache_write":0.04},"canonical_model_id":"amazon/nova-micro"},"mistral.pixtral-large-2502-v1:0":{"id":"mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"openai.gpt-oss-120b":{"id":"openai.gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/v1","shape":"responses"},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"eu.amazon.nova-pro-v1:0":{"id":"eu.amazon.nova-pro-v1:0","name":"Nova Pro (EU)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.92,"output":3.68,"cache_read":0.23,"cache_write":0.92},"canonical_model_id":"amazon/nova-pro"},"global.anthropic.claude-opus-5":{"id":"global.anthropic.claude-opus-5","name":"Claude Opus 5 (Global)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"openai.gpt-6-astra":{"id":"openai.gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":66,"output":330,"cache_read":6.6,"cache_write":82.5},"provider":{"body":{"service_tier":"ultrafast"}}}}},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75,"tiers":[{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5}},"canonical_model_id":"openai/gpt-6-astra"},"global.anthropic.claude-fable-5-1":{"id":"global.anthropic.claude-fable-5-1","name":"Claude Fable 5.1 (Global)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic.claude-fable-5-1":{"id":"anthropic.claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"au.anthropic.claude-opus-5-5":{"id":"au.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (AU)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"jp.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"jp.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (JP)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"moonshot.kimi-k2-thinking":{"id":"moonshot.kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16000},"status":"deprecated","cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"openai.gpt-oss-20b":{"id":"openai.gpt-oss-20b","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/v1","shape":"responses"},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"minimax.minimax-m2.5":{"id":"minimax.minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":98304},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"},"au.anthropic.claude-sonnet-4-6":{"id":"au.anthropic.claude-sonnet-4-6","name":"AU Anthropic Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"global.anthropic.claude-opus-4-6-v1":{"id":"global.anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6 (Global)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"global.xai.grok-4.7":{"id":"global.xai.grok-4.7","name":"Grok 4.7 (Global)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"global.openai.gpt-6.1-sol":{"id":"global.openai.gpt-6.1-sol","name":"GPT-6.1 Sol (Global)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":12,"output":60,"cache_read":0.6,"cache_write":15},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"au.anthropic.claude-opus-4-8":{"id":"au.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (AU)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"global.moonshotai.kimi-k3":{"id":"global.moonshotai.kimi-k3","name":"Kimi K3 (Global)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"moonshotai/kimi-k3"},"jp.amazon.nova-2-lite-v1:0":{"id":"jp.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (JP)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.396,"output":3.311,"cache_read":0.099,"cache_write":0.396},"canonical_model_id":"amazon/nova-2-lite"},"in.anthropic.claude-opus-5":{"id":"in.anthropic.claude-opus-5","name":"Claude Opus 5 (India)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"us.anthropic.claude-fable-5-1":{"id":"us.anthropic.claude-fable-5-1","name":"Claude Fable 5.1 (US)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":0.275,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5-1"},"global.anthropic.claude-opus-4-8":{"id":"global.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (Global)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"au.anthropic.claude-opus-4-6-v1":{"id":"au.anthropic.claude-opus-4-6-v1","name":"AU Anthropic Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"google.gemma-3-27b-it":{"id":"google.gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.23,"output":0.38},"canonical_model_id":"google/gemma-3-27b-it"},"anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"us.anthropic.claude-opus-4-7":{"id":"us.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (US)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"in.openai.gpt-5.6-luna":{"id":"in.openai.gpt-5.6-luna","name":"GPT-5.6 Luna (India)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"anthropic.claude-opus-4-7":{"id":"anthropic.claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"us.anthropic.claude-opus-4-1-20250805-v1:0":{"id":"us.anthropic.claude-opus-4-1-20250805-v1:0","name":"Claude Opus 4.1 (US)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"us.openai.gpt-5.6-luna":{"id":"us.openai.gpt-5.6-luna","name":"GPT-5.6 Luna (US)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"apac.amazon.nova-micro-v1:0":{"id":"apac.amazon.nova-micro-v1:0","name":"Nova Micro (APAC)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.037,"output":0.148,"cache_read":0.00925,"cache_write":0.037},"canonical_model_id":"amazon/nova-micro"},"us.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"us.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (US)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"amazon.nova-micro-v1:0":{"id":"amazon.nova-micro-v1:0","name":"Nova Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875,"cache_write":0.035},"canonical_model_id":"amazon/nova-micro"},"openai.gpt-6.1-sol":{"id":"openai.gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":13.2,"output":66,"cache_read":0.66,"cache_write":16.5},"provider":{"body":{"service_tier":"ultrafast"}}}}},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6.1-sol"},"us.writer.palmyra-x4-v1:0":{"id":"us.writer.palmyra-x4-v1:0","name":"Palmyra X4 (US)","description":"Enterprise language model for workflow automation, coding, data analysis, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-10-09","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":122880,"output":8192},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"writer/palmyra-x4"},"us.anthropic.claude-opus-4-5-20251101-v1:0":{"id":"us.anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5 (US)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"zai.glm-5":{"id":"zai.glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"amazon.nova-2-lite-v1:0":{"id":"amazon.nova-2-lite-v1:0","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.0825,"cache_write":0.33},"canonical_model_id":"amazon/nova-2-lite"},"anthropic.claude-haiku-5-5":{"id":"anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"us.amazon.nova-lite-v1:0":{"id":"us.amazon.nova-lite-v1:0","name":"Nova Lite (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015,"cache_write":0.06},"canonical_model_id":"amazon/nova-lite"},"global.anthropic.claude-haiku-5-5":{"id":"global.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (Global)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"us.writer.palmyra-x5-v1:0":{"id":"us.writer.palmyra-x5-v1:0","name":"Palmyra X5 (US)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"input":1040000,"output":8192},"status":"deprecated","cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"qwen.qwen3-32b-v1:0":{"id":"qwen.qwen3-32b-v1:0","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-32b"},"openai.gpt-6-luna":{"id":"openai.gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275}},"canonical_model_id":"openai/gpt-6-luna"},"us.anthropic.claude-opus-4-6-v1":{"id":"us.anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6 (US)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"global.anthropic.claude-sonnet-5-5":{"id":"global.anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5 (Global)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"google.gemma-3-4b-it":{"id":"google.gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"status":"deprecated","cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"anthropic.claude-opus-4-6-v1":{"id":"anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"us-gov.openai.gpt-oss-20b-1:0":{"id":"us-gov.openai.gpt-oss-20b-1:0","name":"gpt-oss-20b (GovCloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.084,"output":0.36},"canonical_model_id":"openai/gpt-oss-20b"},"openai.gpt-5.6-luna":{"id":"openai.gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"jp.anthropic.claude-opus-4-8":{"id":"jp.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (JP)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"minimax.minimax-m2.1":{"id":"minimax.minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"us.anthropic.claude-opus-5-5":{"id":"us.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (US)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"eu.anthropic.claude-sonnet-5":{"id":"eu.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (EU)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"us.meta.llama3-3-70b-instruct-v1:0":{"id":"us.meta.llama3-3-70b-instruct-v1:0","name":"Llama 3.3 70B Instruct (US)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"in.openai.gpt-5.6-terra":{"id":"in.openai.gpt-5.6-terra","name":"GPT-5.6 Terra (India)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"eu.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"eu.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (EU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"zai.glm-4.7":{"id":"zai.glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"us.anthropic.claude-opus-5":{"id":"us.anthropic.claude-opus-5","name":"Claude Opus 5 (US)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"writer.palmyra-x4-v1:0":{"id":"writer.palmyra-x4-v1:0","name":"Palmyra X4","description":"Enterprise language model for workflow automation, coding, data analysis, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-10-09","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":122880,"output":8192},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"writer/palmyra-x4"},"eu.anthropic.claude-opus-4-7":{"id":"eu.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (EU)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"us.anthropic.claude-fable-5":{"id":"us.anthropic.claude-fable-5","name":"Claude Fable 5 (US)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"openai.gpt-5.5":{"id":"openai.gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-06-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":5.5,"output":33,"cache_read":0.55},"canonical_model_id":"openai/gpt-5.5"},"global.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"global.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (Global)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"us.meta.llama3-1-70b-instruct-v1:0":{"id":"us.meta.llama3-1-70b-instruct-v1:0","name":"Llama 3.1 70B Instruct (US)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"global.anthropic.claude-sonnet-5":{"id":"global.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (Global)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"google.gemma-4-26b-a4b":{"id":"google.gemma-4-26b-a4b","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"apac.amazon.nova-pro-v1:0":{"id":"apac.amazon.nova-pro-v1:0","name":"Nova Pro (APAC)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.84,"output":3.36,"cache_read":0.21,"cache_write":0.84},"canonical_model_id":"amazon/nova-pro"},"us.openai.gpt-6-sol":{"id":"us.openai.gpt-6-sol","name":"GPT-6 Sol (US)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6-sol"},"jp.anthropic.claude-opus-5-5":{"id":"jp.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (JP)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"us-gov.openai.gpt-oss-120b-1:0":{"id":"us-gov.openai.gpt-oss-120b-1:0","name":"gpt-oss-120b (GovCloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.18,"output":0.72},"canonical_model_id":"openai/gpt-oss-120b"},"us.amazon.nova-premier-v1:0":{"id":"us.amazon.nova-premier-v1:0","name":"Nova Premier (US)","description":"Multimodal model for complex analysis, long-context understanding, tool use, and model distillation","family":"nova","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-30","last_updated":"2025-04-30","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":10000},"status":"deprecated","cost":{"input":2.5,"output":12.5,"cache_read":0.625,"cache_write":2.5},"canonical_model_id":"amazon/nova-premier"},"openai.gpt-oss-safeguard-20b":{"id":"openai.gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"us.xai.grok-4.7":{"id":"us.xai.grok-4.7","name":"Grok 4.7 (US)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.2,"output":6.6,"cache_read":0.55},"canonical_model_id":"xai/grok-4.7"},"qwen.qwen3-coder-480b-a35b-v1:0":{"id":"qwen.qwen3-coder-480b-a35b-v1:0","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0.45,"output":1.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"meta.llama4-scout-17b-instruct-v1:0":{"id":"meta.llama4-scout-17b-instruct-v1:0","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10000000,"output":8192},"status":"deprecated","cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"mistral.magistral-small-2509":{"id":"mistral.magistral-small-2509","name":"Magistral Small 1.2","description":"Open multimodal reasoning model for transparent analysis of text and images","family":"magistral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-18","last_updated":"2025-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":40000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/magistral-small-2509"},"openai.gpt-oss-20b-1:0":{"id":"openai.gpt-oss-20b-1:0","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":128000},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"au.anthropic.claude-opus-4-7":{"id":"au.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (AU)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"us.deepseek.r1-v1:0":{"id":"us.deepseek.r1-v1:0","name":"DeepSeek-R1 (US)","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"global.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"global.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (Global)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"us.amazon.nova-2-lite-v1:0":{"id":"us.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (US)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.0825,"cache_write":0.33},"canonical_model_id":"amazon/nova-2-lite"},"qwen.qwen3-vl-235b-a22b":{"id":"qwen.qwen3-vl-235b-a22b","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262000},"cost":{"input":0.53,"output":2.66},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"jp.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"jp.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (JP)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"mistral.ministral-3-3b-instruct":{"id":"mistral.ministral-3-3b-instruct","name":"Ministral 3 3B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"mistral/ministral-3-3b-instruct-2512"},"eu.amazon.nova-lite-v1:0":{"id":"eu.amazon.nova-lite-v1:0","name":"Nova Lite (EU)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.069,"output":0.276,"cache_read":0.01725,"cache_write":0.069},"canonical_model_id":"amazon/nova-lite"},"au.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"au.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (AU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"eu.anthropic.claude-opus-4-6-v1":{"id":"eu.anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6 (EU)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"us.anthropic.claude-sonnet-4-6":{"id":"us.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (US)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"mistral.ministral-3-8b-instruct":{"id":"mistral.ministral-3-8b-instruct","name":"Ministral 3 8B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.15,"output":0.15},"canonical_model_id":"mistral/ministral-3-8b-instruct-2512"},"mistral.devstral-2-123b":{"id":"mistral.devstral-2-123b","name":"Devstral 2 123B","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-2512"},"au.anthropic.claude-opus-5":{"id":"au.anthropic.claude-opus-5","name":"Claude Opus 5 (AU)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"global.openai.gpt-6-sol":{"id":"global.openai.gpt-6-sol","name":"GPT-6 Sol (Global)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"nvidia.nemotron-nano-12b-v2":{"id":"nvidia.nemotron-nano-12b-v2","name":"NVIDIA Nemotron Nano 12B v2 VL BF16","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.2,"output":0.6},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"jp.anthropic.claude-opus-5":{"id":"jp.anthropic.claude-opus-5","name":"Claude Opus 5 (JP)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"us.anthropic.claude-opus-4-8":{"id":"us.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (US)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"global.xai.grok-4.6":{"id":"global.xai.grok-4.6","name":"Grok 4.6 (Global)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"us.openai.gpt-6-astra":{"id":"us.openai.gpt-6-astra","name":"GPT-6 Astra (US)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":66,"output":330,"cache_read":6.6,"cache_write":82.5},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75,"tiers":[{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5}},"canonical_model_id":"openai/gpt-6-astra"},"global.anthropic.claude-opus-5-5":{"id":"global.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (Global)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"us.openai.gpt-6.1-sol":{"id":"us.openai.gpt-6.1-sol","name":"GPT-6.1 Sol (US)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":13.2,"output":66,"cache_read":0.66,"cache_write":16.5},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai.gpt-oss-120b-1:0":{"id":"openai.gpt-oss-120b-1:0","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":128000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"openai.gpt-5.4":{"id":"openai.gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-06-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.75,"output":16.5,"cache_read":0.275},"canonical_model_id":"openai/gpt-5.4"},"eu.anthropic.claude-opus-4-5-20251101-v1:0":{"id":"eu.anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5 (EU)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic.claude-opus-4-1-20250805-v1:0":{"id":"anthropic.claude-opus-4-1-20250805-v1:0","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"global.amazon.nova-2-lite-v1:0":{"id":"global.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (Global)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.075,"cache_write":0.3},"canonical_model_id":"amazon/nova-2-lite"},"global.openai.gpt-5.6-terra":{"id":"global.openai.gpt-5.6-terra","name":"GPT-5.6 Terra (Global)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai.gpt-oss-safeguard-120b":{"id":"openai.gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"openai.gpt-5.6-terra":{"id":"openai.gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"zai.glm-4.7-flash":{"id":"zai.glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"xai.grok-4.6":{"id":"xai.grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":6.6,"cache_read":0.55},"canonical_model_id":"xai/grok-4.6"},"global.openai.gpt-6-astra":{"id":"global.openai.gpt-6-astra","name":"GPT-6 Astra (Global)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":60,"output":300,"cache_read":6,"cache_write":75},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"us.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"us.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (US)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"us.openai.gpt-6-luna":{"id":"us.openai.gpt-6-luna","name":"GPT-6 Luna (US)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275}},"canonical_model_id":"openai/gpt-6-luna"},"eu.anthropic.claude-opus-5":{"id":"eu.anthropic.claude-opus-5","name":"Claude Opus 5 (EU)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"eu.anthropic.claude-opus-5-5":{"id":"eu.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (EU)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"mistral.voxtral-mini-3b-2507":{"id":"mistral.voxtral-mini-3b-2507","name":"Voxtral Mini 3B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"status":"deprecated","cost":{"input":0.04,"output":0.04},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"eu.anthropic.claude-sonnet-4-6":{"id":"eu.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (EU)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic.claude-opus-5-5":{"id":"anthropic.claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"us.openai.gpt-5.6-sol":{"id":"us.openai.gpt-5.6-sol","name":"GPT-5.6 Sol (US)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5,"tiers":[{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11}},"canonical_model_id":"openai/gpt-5.6-sol"},"global.zai.glm-5.3":{"id":"global.zai.glm-5.3","name":"GLM-5.3 (Global)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.68,"output":5.28,"cache_read":0.312,"cache_write":2.1},"canonical_model_id":"zhipuai/glm-5.3"},"us.mistral.pixtral-large-2502-v1:0":{"id":"us.mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02) (US)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"jp.anthropic.claude-sonnet-4-6":{"id":"jp.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (JP)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"au.anthropic.claude-sonnet-5":{"id":"au.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (AU)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"jp.anthropic.claude-sonnet-5":{"id":"jp.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (JP)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"us.moonshotai.kimi-k3":{"id":"us.moonshotai.kimi-k3","name":"Kimi K3 (US)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"moonshotai/kimi-k3"},"anthropic.claude-opus-4-8":{"id":"anthropic.claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"google.gemma-4-e2b":{"id":"google.gemma-4-e2b","name":"Gemma 4 E2B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-4-E2B-it"},"google.gemma-4-31b":{"id":"google.gemma-4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"anthropic.claude-sonnet-5-5":{"id":"anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"qwen.qwen3-coder-30b-a3b-v1:0":{"id":"qwen.qwen3-coder-30b-a3b-v1:0","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"mistral.mistral-large-3-675b-instruct":{"id":"mistral.mistral-large-3-675b-instruct","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"nvidia.nemotron-super-3-120b":{"id":"nvidia.nemotron-super-3-120b","name":"NVIDIA Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.15,"output":0.65},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"jp.anthropic.claude-opus-4-7":{"id":"jp.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (JP)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"global.anthropic.claude-opus-4-5-20251101-v1:0":{"id":"global.anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5 (Global)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"apac.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"apac.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (APAC)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"eu.anthropic.claude-fable-5":{"id":"eu.anthropic.claude-fable-5","name":"Claude Fable 5 (EU)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"deepseek.v3.2":{"id":"deepseek.v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2026-02-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":81920},"cost":{"input":0.62,"output":1.85},"canonical_model_id":"deepseek/deepseek-v3.2"},"us.meta.llama4-scout-17b-instruct-v1:0":{"id":"us.meta.llama4-scout-17b-instruct-v1:0","name":"Llama 4 Scout 17B Instruct (US)","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10000000,"output":8192},"status":"deprecated","cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"anthropic.claude-opus-4-5-20251101-v1:0":{"id":"anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"mistral.voxtral-small-24b-2507":{"id":"mistral.voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"status":"deprecated","cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"eu.anthropic.claude-haiku-5-5":{"id":"eu.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (EU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"global.openai.gpt-6-luna":{"id":"global.openai.gpt-6-luna","name":"GPT-6 Luna (Global)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"global.openai.gpt-5.6-sol":{"id":"global.openai.gpt-5.6-sol","name":"GPT-5.6 Sol (Global)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"global.anthropic.claude-opus-4-7":{"id":"global.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (Global)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"in.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"in.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (India)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"}}},"minimax-cn":{"id":"minimax-cn","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.cn/anthropic/v1","name":"MiniMax (minimax.cn)","doc":"https://platform.minimaxi.com/docs/guides/quickstart","models":{"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6275,"output":2.5098,"cache_read":0.0314,"cache_write":0.3922}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0627,"cache_write":0.3922}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0314,"cache_write":0.3922}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0.3137,"output":1.2549,"cache_read":0.0627,"tiers":[{"input":0.6275,"output":2.5098,"cache_read":0.1255,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6275,"output":2.5098,"cache_read":0.1255}}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0314,"cache_write":0.3922}},"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6275,"output":2.5098,"cache_read":0.0627,"cache_write":0.3922}},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0314,"cache_write":0.3922}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.1494,"output":0.5976,"cache_read":0.0149},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"}}},"cerebras":{"id":"cerebras","env":["CEREBRAS_API_KEY"],"npm":"@ai-sdk/cerebras","name":"Cerebras","doc":"https://inference-docs.cerebras.ai/models/overview","models":{"qwen-3.8-27b":{"id":"qwen-3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.99,"output":1.49,"cache_read":0.99},"canonical_model_id":"alibaba/qwen3.8-27b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.35,"output":0.75,"cache_read":0.35},"canonical_model_id":"openai/gpt-oss-120b"}}},"pioneer":{"id":"pioneer","env":["PIONEER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.fastino.ai/v1","name":"Pioneer","doc":"https://agent.pioneer.ai/llms.txt","models":{"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-terra"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":1,"output":6,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"openai/gpt-5.6-luna"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-5-fast":{"id":"claude-opus-5-fast","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"status":"deprecated","cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5},"canonical_model_id":"openai/gpt-5.5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"openai/gpt-5.6-sol"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.1,"output":0.2,"cache_read":0.0197,"cache_write":0.1},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K3-Fast":{"id":"moonshotai/Kimi-K3-Fast","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":4.5,"output":22.5,"cache_read":0.45,"cache_write":4.5},"canonical_model_id":"moonshotai/kimi-k3"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"cost":{"input":0.08,"output":0.45,"cache_read":0.04,"cache_write":0.08},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":4096},"cost":{"input":0.07,"output":0.2,"cache_read":0.035,"cache_write":0.07},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"fastino/GLiNER-2.5-Decide":{"id":"fastino/GLiNER-2.5-Decide","name":"GLiNER-2.5-Decide","description":"Classification-tuned GLiNER2 large. Entities, relations, and JSON extraction still work.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-09-24","last_updated":"2026-09-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-privacy-filter-PII-multi":{"id":"fastino/gliner2-privacy-filter-PII-multi","name":"GLiNER2 Privacy Filter PII (Multi)","description":"Multilingual PII entity detection across 40+ languages; non-trainable encoder.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-multi-v1":{"id":"fastino/gliner2-multi-v1","name":"GLiNER2 Multi","description":"Multilingual NER for cross-lingual extraction.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-11-30","last_updated":"2025-11-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2.5-base-v1":{"id":"fastino/gliner2.5-base-v1","name":"GLiNER 2.5 Base","description":"English boundary NER and structured extraction with LoRA fine-tuning.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-08-24","last_updated":"2026-08-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-large-v1":{"id":"fastino/gliner2-large-v1","name":"GLiNER2 Large","description":"Higher-accuracy NER; broader entity coverage.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2.5-multi-v1":{"id":"fastino/gliner2.5-multi-v1","name":"GLiNER 2.5 Multi","description":"Multilingual boundary NER and span extraction; non-trainable encoder.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-08-24","last_updated":"2026-08-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-base-v1":{"id":"fastino/gliner2-base-v1","name":"GLiNER2 Base","description":"Named entity recognition; zero-shot span extraction.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":2048,"output":2048},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliguard-PII-multi":{"id":"fastino/gliguard-PII-multi","name":"GLiNER2-Guardrails-PII-Multi","description":"A 300M-parameter multilingual model that runs LLM safety moderation and PII detection in a single forward pass.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.15,"output":0.15,"cache_read":0.15,"cache_write":0.15}},"fastino/gliguard-LLMGuardrails-300M":{"id":"fastino/gliguard-LLMGuardrails-300M","name":"GLiGuard LLM Guardrails 300M","description":"LLM safety, PII, and adversarial-attack guardrails; 300M-param encoder.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.15,"output":0.15,"cache_read":0.15,"cache_write":0.15}},"fastino/gliner2-multi-large-v1":{"id":"fastino/gliner2-multi-large-v1","name":"GLiNER2 Multi Large","description":"High-accuracy multilingual NER and span extraction.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-11-30","last_updated":"2025-11-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1040000,"output":128000},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":1.4},"canonical_model_id":"zhipuai/glm-5.2"}}},"cloudferro-sherlock":{"id":"cloudferro-sherlock","env":["CLOUDFERRO_SHERLOCK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api-sherlock.cloudferro.com/openai/v1/","name":"CloudFerro Sherlock","doc":"https://docs.sherlock.cloudferro.com/","models":{"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"input":180000,"output":16000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"},"speakleash/Bielik-11B-v3.0-Instruct":{"id":"speakleash/Bielik-11B-v3.0-Instruct","name":"Bielik 11B v3.0 Instruct","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.67,"output":0.67}},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"OpenAI GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":2.92,"output":2.92},"canonical_model_id":"openai/gpt-oss-120b"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10-09","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":70000,"output":70000},"cost":{"input":2.92,"output":2.92},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.34,"output":4.48},"canonical_model_id":"zhipuai/glm-5.3"}}},"reflection":{"id":"reflection","env":["REFLECTION_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.reflection.ai/openai/v1","name":"Reflection","doc":"https://developers.reflection.ai/models","models":{"Beam-1.0-501B-A23B":{"id":"Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"}}},"llmtech":{"id":"llmtech","env":["LLMTECH_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.llmtech.eu/v1","name":"LLM Tech","doc":"https://llmtech.eu/models/qwen3.8-27b","models":{"nvidia/Qwen3.8-27B-NVFP4":{"id":"nvidia/Qwen3.8-27B-NVFP4","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.25,"output":2.09,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"zai":{"id":"zai","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.z.ai/api/paas/v4","name":"Z.AI","doc":"https://docs.z.ai/guides/overview/pricing","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"glm-5-turbo":{"id":"glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5.3-flashx":{"id":"glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5v"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"glm-4.7-flashx":{"id":"glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"glm-4.6v-flash":{"id":"glm-4.6v-flash","name":"GLM-4.6V-Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5v-turbo"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"glm-4.5-flash":{"id":"glm-4.5-flash","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-flash"}}},"sensenova":{"id":"sensenova","env":["SENSENOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token.sensenova.cn/v1","name":"SenseNova (China)","doc":"https://platform.sensenova.cn/docs","models":{"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash is a high-performance conversational model by DeepSeek, supporting thinking/non-thinking modes, 1M context window, and tool calling.","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.2"},"sensenova-6.8-flash-lite":{"id":"sensenova-6.8-flash-lite","name":"SenseNova 6.8 Flash Lite","description":"SenseNova lightweight multimodal agent model for real-world complex tasks, data analysis, and complex information presentation","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-11","last_updated":"2026-08-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0,"cache_read":0}}}},"lucidquery":{"id":"lucidquery","env":["LUCIDQUERY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lucidquery.com/v1","name":"LucidQuery","doc":"https://lucidquery.com/docs","models":{"lucidquery-nexus-coder":{"id":"lucidquery-nexus-coder","name":"LucidQuery Nexus Coder","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"lucid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-08-01","release_date":"2025-09-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":250000,"output":60000},"cost":{"input":2,"output":5}},"lucidnova-rf1-100b":{"id":"lucidnova-rf1-100b","name":"LucidNova RF1 100B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"nova","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-09-16","release_date":"2024-12-28","last_updated":"2025-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":120000,"output":8000},"cost":{"input":2,"output":5}},"lucidquery-agi-01-frontier":{"id":"lucidquery-agi-01-frontier","name":"AGI-01 Frontier","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-06-05","release_date":"2026-06-16","last_updated":"2026-06-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":120000},"cost":{"input":5.04675,"output":24.673}},"lucidquery-agi-01-swift":{"id":"lucidquery-agi-01-swift","name":"AGI-01 Swift","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-06-05","release_date":"2026-06-16","last_updated":"2026-06-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":120000},"cost":{"input":2.80375,"output":16.8225}}}},"kimi-code-plan-global":{"id":"kimi-code-plan-global","env":["KIMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.kimi.ai/coding/v1","name":"Kimi For Coding (kimi.ai)","doc":"https://www.kimi.ai/code/docs/en/kimi-code/models.html","models":{"k3":{"id":"k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"k3.1-pro":{"id":"k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-for-coding":{"id":"kimi-for-coding","name":"kimi-for-coding","description":"Kimi coding model with more efficient thinking and up to 1M context, available through Kimi Code","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.8-preview"},"k3-256k":{"id":"k3-256k","name":"Kimi K3-256K","description":"256K-context version of Kimi K3, reducing token consumption for shorter coding sessions","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-for-coding-highspeed":{"id":"kimi-for-coding-highspeed","name":"Kimi For Coding HighSpeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"k3.1-pro-highspeed":{"id":"k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"}}},"alibaba":{"id":"alibaba","env":["DASHSCOPE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://dashscope-intl.aliyuncs.com/compatible-mode/v1","name":"Alibaba","doc":"https://www.alibabacloud.com/help/en/model-studio/models","models":{"qwen3-livetranslate-flash-realtime":{"id":"qwen3-livetranslate-flash-realtime","type":"realtime","name":"Qwen3-LiveTranslate Flash Realtime","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":53248,"output":4096},"status":"deprecated","provider":{"api":"wss://dashscope-intl.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":10,"output":10,"input_audio":10,"output_audio":38}},"qwen-vl-max":{"id":"qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.8,"output":3.2}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"qwen2-5-omni-7b":{"id":"qwen2-5-omni-7b","name":"Qwen2.5-Omni 7B","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-12","last_updated":"2024-12","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":true,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.1,"output":0.4,"input_audio":6.76}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}}},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen3 235B-A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":2.8,"reasoning":8.4}},"qwen-mt-plus":{"id":"qwen-mt-plus","name":"Qwen-MT Plus","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":2.46,"output":7.37}},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.45,"output":2.25,"tiers":[{"input":0.75,"output":3.75,"tier":{"type":"context","size":32000}},{"input":1.2,"output":6,"tier":{"type":"context","size":128000}}]}},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":2.8,"reasoning":8.4}},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":131072},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.0375,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen-omni-turbo-realtime":{"id":"qwen-omni-turbo-realtime","type":"realtime","name":"Qwen-Omni Turbo Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-05-08","last_updated":"2025-05-08","modalities":{"input":["text","image","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","provider":{"api":"wss://dashscope-intl.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.27,"output":1.07,"input_audio":4.44,"output_audio":8.89}},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"Qwen3-VL 235B-A22B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.7,"output":2.8,"reasoning":8.4}},"qwen3-coder-flash":{"id":"qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.1,"cache_write":0.625,"tier":{"type":"context","size":32000}},{"input":0.8,"output":4,"cache_read":0.16,"cache_write":1,"tier":{"type":"context","size":128000}},{"input":1.6,"output":9.6,"cache_read":0.32,"cache_write":2,"tier":{"type":"context","size":256000}}]}},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125}},"qwen3-next-80b-a3b-thinking":{"id":"qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":6}},"qwen3-8b":{"id":"qwen3-8b","name":"Qwen3 8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.18,"output":0.7,"reasoning":2.1}},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2}},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":1.6,"cache_read":0.04,"cache_write":0.25,"tiers":[{"input":0.3,"output":2.4,"cache_read":0.06,"cache_write":0.375,"tier":{"type":"context","size":32000}},{"input":0.6,"output":4.8,"cache_read":0.12,"cache_write":0.75,"tier":{"type":"context","size":128000}}]}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1,"output":5}},"qwen-vl-ocr":{"id":"qwen-vl-ocr","name":"Qwen-VL OCR","description":"OCR model for extracting structured text from documents and screenshots","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-10-28","last_updated":"2025-04-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":34096,"output":4096},"cost":{"input":0.07,"output":0.16}},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.2,"output":6}},"qwen-flash":{"id":"qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.05,"output":0.4,"cache_read":0.01,"cache_write":0.0625,"tiers":[{"input":0.25,"output":2,"cache_read":0.05,"cache_write":0.3125,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.25,"output":2,"cache_read":0.05,"cache_write":0.3125}}},"qwen3.8-omni-flash":{"id":"qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen3.5-122b-a10b":{"id":"qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2}},"qwen3-omni-flash-realtime":{"id":"qwen3-omni-flash-realtime","type":"realtime","name":"Qwen3-Omni Flash Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"status":"deprecated","provider":{"api":"wss://dashscope-intl.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.52,"output":1.99,"input_audio":4.57,"output_audio":18.13}},"qwen3.6-max-preview":{"id":"qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.3,"output":7.8,"cache_read":0.13,"cache_write":1.625}},"qwen-turbo":{"id":"qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":16384},"status":"deprecated","cost":{"input":0.05,"output":0.2,"reasoning":0.5}},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6}},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.088},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qvq-max":{"id":"qvq-max","name":"QVQ Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qvq","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":1.2,"output":4.8}},"qwen-mt-turbo":{"id":"qwen-mt-turbo","name":"Qwen-MT Turbo","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"status":"deprecated","cost":{"input":0.16,"output":0.49}},"qwen3-14b":{"id":"qwen3-14b","name":"Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.35,"output":1.4,"reasoning":4.2}},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"reasoning":2.4,"tiers":[{"input":0.5,"output":3,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.5,"output":3}}},"qwen3-vl-30b-a3b":{"id":"qwen3-vl-30b-a3b","name":"Qwen3-VL 30B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.8,"reasoning":2.4}},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6}},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.28,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3-asr-flash":{"id":"qwen3-asr-flash","type":"transcription","name":"Qwen3-ASR Flash","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-04","release_date":"2025-09-08","last_updated":"2025-09-08","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":53248,"output":4096}},"qwen-plus":{"id":"qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.2,"reasoning":4,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":3.6,"reasoning":12,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":3.6,"reasoning":12,"cache_read":0.24,"cache_write":1.5}}},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5}}},"qwen-max":{"id":"qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4}},"qwen-vl-plus":{"id":"qwen-vl-plus","name":"Qwen-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.21,"output":0.63}},"qwen-plus-character-ja":{"id":"qwen-plus-character-ja","name":"Qwen Plus Character (Japanese)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":512},"cost":{"input":0.5,"output":1.4}},"qwen-omni-turbo":{"id":"qwen-omni-turbo","name":"Qwen-Omni Turbo","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-19","last_updated":"2025-03-26","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.07,"output":0.27,"input_audio":4.44,"output_audio":8.89}},"qwen3.5-flash":{"id":"qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.5,"output":7.5,"tiers":[{"input":2.7,"output":13.5,"tier":{"type":"context","size":32000}},{"input":4.5,"output":22.5,"tier":{"type":"context","size":128000}}]}},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4}},"qwen3-omni-flash":{"id":"qwen3-omni-flash","name":"Qwen3-Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"cost":{"input":0.43,"output":1.66,"input_audio":3.81,"output_audio":15.11}},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":2}},"qwq-plus":{"id":"qwq-plus","name":"QwQ Plus","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.8,"output":2.4}},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.375,"output":2.25}},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.3125,"tiers":[{"input":1,"output":4,"cache_read":0.1,"cache_write":1.25,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1,"output":4,"cache_read":0.1,"cache_write":1.25}}}}},"watsonx":{"id":"watsonx","env":["WATSONX_AI_APIKEY","WATSONX_AI_PROJECT_ID"],"npm":"watsonx-ai-provider","name":"watsonx.ai","doc":"https://www.ibm.com/docs/en/watsonx/saas?topic=solutions-supported-foundation-models","models":{"ibm/granite-4-h-small":{"id":"ibm/granite-4-h-small","name":"Granite-4.0-H-Small","description":"Open-weight hybrid model for enterprise chat, coding, retrieval-augmented generation, and tool-calling workloads","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0636,"output":0.265},"canonical_model_id":"ibm/granite-4-h-small"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.159,"output":0.636},"canonical_model_id":"openai/gpt-oss-120b"},"mistralai/mistral-small-3-1-24b-instruct-2503":{"id":"mistralai/mistral-small-3-1-24b-instruct-2503","name":"Mistral Small 3.1 24B","description":"Efficient multimodal model for instruction following, coding, reasoning, and function calling","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.106,"output":0.318},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"meta-llama/llama-3-3-70b-instruct":{"id":"meta-llama/llama-3-3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.7526,"output":0.7526},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-4-maverick-17b-128e-instruct-fp8":{"id":"meta-llama/llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B 128E Instruct FP8","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.371,"output":1.484},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"}}},"groq":{"id":"groq","env":["GROQ_API_KEY"],"npm":"@ai-sdk/groq","name":"Groq","doc":"https://console.groq.com/docs/models","models":{"llama-3.3-70b-versatile":{"id":"llama-3.3-70b-versatile","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"whisper-large-v3":{"id":"whisper-large-v3","type":"transcription","name":"Whisper","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"canonical_model_id":"openai/whisper-large-v3"},"llama-3.1-8b-instant":{"id":"llama-3.1-8b-instant","name":"Llama 3.1 8B","description":"Compact Llama instruction model for fast chat and local deployment","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.05,"output":0.08},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"whisper-large-v3-turbo":{"id":"whisper-large-v3-turbo","type":"transcription","name":"Whisper Large V3 Turbo","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"canonical_model_id":"openai/whisper-large-v3-turbo"},"allam-2-7b":{"id":"allam-2-7b","name":"ALLaM-2-7b","description":"ALLaM-2-7b instruction tuned model by SDAIA","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-01-23","last_updated":"2025-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"sdaia/allam-2-7b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"Safety GPT OSS 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"beta","cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","default","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"beta","cost":{"input":0.8,"output":4},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","default"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.6,"output":3,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.6-27b"},"meta-llama/llama-prompt-guard-2-86m":{"id":"meta-llama/llama-prompt-guard-2-86m","name":"Prompt Guard 2 86M","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"status":"beta","cost":{"input":0.04,"output":0.04},"canonical_model_id":"meta/llama-prompt-guard-2-86m"},"meta-llama/llama-prompt-guard-2-22m":{"id":"meta-llama/llama-prompt-guard-2-22m","name":"Llama Prompt Guard 2 22M","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"status":"beta","cost":{"input":0.03,"output":0.03},"canonical_model_id":"meta/llama-prompt-guard-2-22m"},"canopylabs/orpheus-arabic-saudi":{"id":"canopylabs/orpheus-arabic-saudi","type":"speech","name":"Canopy Labs Orpheus Arabic Saudi","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"canopylabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":4000,"output":50000},"status":"beta","provider":{"npm":"@ai-sdk/openai","api":"https://api.groq.com/openai/v1"},"canonical_model_id":"canopylabs/orpheus-arabic-saudi"},"canopylabs/orpheus-v1-english":{"id":"canopylabs/orpheus-v1-english","type":"speech","name":"Canopy Labs Orpheus V1 English","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"canopylabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-19","last_updated":"2025-12-19","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":4000,"output":50000},"status":"beta","provider":{"npm":"@ai-sdk/openai","api":"https://api.groq.com/openai/v1"},"canonical_model_id":"canopylabs/orpheus-v1-english"}}},"modelis":{"id":"modelis","env":["MODELIS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://modelishub.com/v1","name":"Modelis","doc":"https://modelishub.com/pricing","models":{"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.435,"output":0.87},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10},"canonical_model_id":"google/gemini-2.5-pro"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.0983,"output":0.1966},"canonical_model_id":"deepseek/deepseek-v4-flash"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":9},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.768,"output":3.072},"canonical_model_id":"alibaba/qwen3.7-plus"}}},"crossmodel":{"id":"crossmodel","env":["CROSSMODEL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.crossmodel.ai/v1","name":"CrossModel","doc":"https://www.crossmodel.ai/docs","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.35,"output":4.05,"cache_read":0.045,"cache_write":1.35},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"cache_write":1.25,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"cache_write":1,"tiers":[{"input":2,"output":4,"cache_read":0.4,"cache_write":2,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4,"cache_write":2}},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":0.6,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6,"cache_write":4}},"canonical_model_id":"xai/grok-4.5"},"gemini/gemini-2.5-pro":{"id":"gemini/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":1.25,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5}},"canonical_model_id":"google/gemini-2.5-pro"},"gemini/gemini-3.6-flash":{"id":"gemini/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"gemini/gemini-3.5-flash-lite":{"id":"gemini/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gemini/gemini-4-argon":{"id":"gemini/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2},"canonical_model_id":"google/gemini-4-argon"},"gemini/gemini-3.1-pro-preview":{"id":"gemini/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini/gemini-3.7-flash":{"id":"gemini/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"gemini/gemini-3-flash-preview":{"id":"gemini/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.5},"canonical_model_id":"google/gemini-3-flash-preview"},"gemini/gemini-2.5-flash":{"id":"gemini/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"google/gemini-2.5-flash"},"gemini/gemini-3.8-flash":{"id":"gemini/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"gemini/gemini-3.5-flash":{"id":"gemini/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gemini/gemini-2.5-flash-lite":{"id":"gemini/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.1},"canonical_model_id":"google/gemini-2.5-flash-lite"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.33,"output":1.32,"cache_read":0.066,"cache_write":0.42},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":512000},"cost":{"input":0.33,"output":1.32,"cache_read":0.066,"cache_write":0.33,"tiers":[{"input":0.66,"output":2.63,"cache_read":0.132,"cache_write":0.66,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.66,"output":2.63,"cache_read":0.132,"cache_write":0.66}},"canonical_model_id":"minimax/MiniMax-M3"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.16,"output":0.32,"cache_read":0.004,"cache_write":0.16},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.47,"output":0.94,"cache_read":0.005,"cache_write":0.47},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5,"cache_write":5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075,"cache_write":0.15},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02,"cache_write":0.2},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":10}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"openai/gpt-5.4-mini"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":1,"output":4.16,"cache_read":0.18,"cache_write":1},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":1,"output":4.16,"cache_read":0.18,"cache_write":1},"canonical_model_id":"moonshotai/kimi-k2.6"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.32,"output":1.88,"cache_read":0.032,"cache_write":0.4,"tiers":[{"input":1.25,"output":7.5,"cache_read":0.124,"cache_write":1.57,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.25,"output":7.5,"cache_read":0.124,"cache_write":1.57}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.04,"output":0.13,"cache_read":0.01,"cache_write":0.04,"tiers":[{"input":0.1,"output":0.37,"cache_read":0.02,"cache_write":0.12,"tier":{"type":"context","size":32000}},{"input":0.19,"output":0.74,"cache_read":0.04,"cache_write":0.24,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.88,"output":5.63,"cache_read":0.375,"cache_write":2.35},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.13,"output":0.43,"cache_read":0.016,"cache_write":0.13},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.88,"output":5.63,"cache_read":0.23,"cache_write":2.35},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.13,"output":0.43,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.32,"output":1.25,"cache_read":0.032,"cache_write":0.4,"tiers":[{"input":0.96,"output":3.75,"cache_read":0.096,"cache_write":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.96,"output":3.75,"cache_read":0.096,"cache_write":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.19,"output":1.13,"cache_read":0.019,"cache_write":0.24,"tiers":[{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.94,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.94}},"canonical_model_id":"alibaba/qwen3.6-flash"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":131072},"cost":{"input":0.16,"output":0.64,"cache_read":0.04,"cache_write":0.16},"canonical_model_id":"tencent/hy3"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0.96,"output":2.88,"cache_read":0.048,"cache_write":0.96},"canonical_model_id":"tencent/hy4-preview"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":3,"cache_read":0.16,"cache_write":0.6,"tiers":[{"input":0.8,"output":3.4,"cache_read":0.2,"cache_write":0.8,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.9,"output":3.7,"cache_read":0.18,"cache_write":0.9,"tiers":[{"input":1.1,"output":4.3,"cache_read":0.27,"cache_write":1.1,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.2,"output":4.4,"cache_read":0.3,"cache_write":1.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1,"output":3.8,"cache_read":0.2,"cache_write":1,"tiers":[{"input":1.2,"output":4.4,"cache_read":0.3,"cache_write":1.2,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.47,"output":2.16,"cache_read":0.1,"cache_write":0.47,"tiers":[{"input":0.62,"output":2.47,"cache_read":0.13,"cache_write":0.62,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.2,"output":4.4,"cache_read":0.3,"cache_write":1.2},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0.15},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"aki-io":{"id":"aki-io","env":["AKI_IO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://aki.io/v1","name":"AKI.IO","doc":"https://aki.io/docs/","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.33645,"output":2.4673,"cache_read":0.11215},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-v4.1-flash-552b":{"id":"deepseek-v4.1-flash-552b","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":81920},"cost":{"input":0.280375,"output":1.1215,"cache_read":0.11215},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm5.3-754b":{"id":"glm5.3-754b","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":81920},"cost":{"input":1.1215,"output":3.92525,"cache_read":0.280375},"canonical_model_id":"zhipuai/glm-5.3"},"mistral4-119b":{"id":"mistral4-119b","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.2243,"output":0.6729},"canonical_model_id":"mistral/mistral-small-2603"},"gemma4-26b":{"id":"gemma4-26b","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0.11215,"output":0.56075},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.168225,"output":0.616825},"canonical_model_id":"openai/gpt-oss-120b"},"qwen3.6-35b":{"id":"qwen3.6-35b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"status":"deprecated","cost":{"input":0.168225,"output":0.56075},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"trustedrouter":{"id":"trustedrouter","env":["TRUSTEDROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.trustedrouter.com/v1","name":"TrustedRouter","doc":"https://trustedrouter.com/docs","models":{"trustedrouter/synth-code":{"id":"trustedrouter/synth-code","name":"Synth Code","description":"TrustedRouter code synthesis orchestration alias that combines multiple model responses into one answer.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-20","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/e2e":{"id":"trustedrouter/e2e","name":"End-to-End Encrypted","description":"TrustedRouter privacy routing alias for end-to-end encrypted provider routes where available.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/auto":{"id":"trustedrouter/auto","name":"Auto","description":"TrustedRouter automatic routing alias that chooses a healthy supported model endpoint for the request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/synth":{"id":"trustedrouter/synth","name":"Synth","description":"TrustedRouter synthesis orchestration alias that combines multiple model responses into one answer.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-20","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/cheap":{"id":"trustedrouter/cheap","name":"Cheap","description":"TrustedRouter low-cost routing alias that prefers inexpensive healthy model endpoints.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/zdr":{"id":"trustedrouter/zdr","name":"Zero Data Retention","description":"TrustedRouter privacy routing alias that prefers zero data retention model endpoints.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/fast":{"id":"trustedrouter/fast","name":"Fast","description":"TrustedRouter speed routing alias that prefers low-latency healthy model endpoints.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}}}},"jalapeno":{"id":"jalapeno","env":["JALAPENO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.jalapeno-cloud.ai/v1","name":"Jalapeno Cloud","doc":"https://www.jalapeno-cloud.ai/docs/","models":{"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.6,"output":3.38,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v4-pro"},"Kimi-K2.5":{"id":"Kimi-K2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":180224},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"Kimi-K3":{"id":"Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"Qwen3-VL-235B-A22B-Thinking":{"id":"Qwen3-VL-235B-A22B-Thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"Qwen3.5-35B-A3B":{"id":"Qwen3.5-35B-A3B","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"Qwen3.5-27B":{"id":"Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen3-VL-235B-A22B-Instruct":{"id":"Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":129024,"output":32768},"status":"deprecated","cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"Kimi-K2.7-Code":{"id":"Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":271360,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":129024,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"Qwen3-Next-80B-A3B-Thinking":{"id":"Qwen3-Next-80B-A3B-Thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"Hy3":{"id":"Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"input":192000,"output":128000},"status":"deprecated","cost":{"input":0.14,"output":0.58},"canonical_model_id":"tencent/hy3"},"Qwen3.5-397B-A17B":{"id":"Qwen3.5-397B-A17B","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen3.5-122B-A10B":{"id":"Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"}}},"nebius":{"id":"nebius","env":["NEBIUS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tokenfactory.nebius.com/v1","name":"Nebius Token Factory","doc":"https://docs.tokenfactory.nebius.com/","models":{"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3-30B-A3B-Instruct-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":8192},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.2,"output":0.6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3-Embedding-8B":{"id":"Qwen/Qwen3-Embedding-8B","type":"embedding","name":"Qwen3-Embedding-8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-10","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"input":40960,"output":0},"cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5-397B-A17B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":250000,"output":8192},"cost":{"input":0.6,"output":3.6,"cache_read":0.12},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"reflection/beam":{"id":"reflection/beam","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.25,"output":1,"cache_read":0.05},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma-3-27b-it","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":110000,"input":100000,"output":8192},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"google/gemma-3-27b-it"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"gpt-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":124000,"output":8192},"cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.15},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":1024000},"cost":{"input":0.14,"output":0.28,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.75,"output":3.5,"cache_read":1.75},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":979000,"output":979000},"cost":{"input":1.32,"output":3.96,"cache_read":0.04},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"output":1048000},"cost":{"input":0.3,"output":1.2,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"meta-models/muse-spark-open":{"id":"meta-models/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":8000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8000},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"NousResearch/Hermes-4-405B":{"id":"NousResearch/Hermes-4-405B","name":"Hermes-4-405B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":120000,"output":8192},"cost":{"input":1,"output":3,"reasoning":3,"cache_read":1},"canonical_model_id":"nousresearch/hermes-4-405b"},"nvidia/Nemotron-3.5-Super-VL":{"id":"nvidia/Nemotron-3.5-Super-VL","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for coding, visual understanding, and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/Nemotron-3-Ultra-550b-a55b":{"id":"nvidia/Nemotron-3-Ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":3,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/Nemotron-3_5-Lightning":{"id":"nvidia/Nemotron-3_5-Lightning","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.06,"output":0.24,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron-3-Super-120B-A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":1024000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":1024000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"}}},"stepfun-step-plan":{"id":"stepfun-step-plan","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.com/step_plan/v1","name":"StepFun Step Plan (China)","doc":"https://platform.stepfun.com/docs/zh/step-plan/integrations/reasoning-api","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.7-flash"},"step-router-v1":{"id":"step-router-v1","name":"Step Router v1","description":"StepFun routing model that dispatches requests to the appropriate Step model.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":256000}},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash"}}},"thinkingmachines":{"id":"thinkingmachines","env":["TINKER_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://tinker.thinkingmachines.dev/services/tinker-prod/anthropic/api/v1","name":"Thinking Machines","doc":"https://tinker-docs.thinkingmachines.ai/tinker/compatible-apis/anthropic/","models":{"Qwen/Qwen3.5-4B":{"id":"Qwen/Qwen3.5-4B","name":"Qwen3.5 4B","description":"Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.33,"output":1.005,"cache_read":0.066},"canonical_model_id":"alibaba/qwen3.5-4b"},"Qwen/Qwen3.5-397B-A17B:peft:262144":{"id":"Qwen/Qwen3.5-397B-A17B:peft:262144","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.65,"output":7.5,"cache_read":0.33},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3.5-4B:peft:262144":{"id":"Qwen/Qwen3.5-4B:peft:262144","name":"Qwen3.5 4B","description":"Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.33,"output":1.005,"cache_read":0.066},"canonical_model_id":"alibaba/qwen3.5-4b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.86,"output":5.595,"cache_read":0.372},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3.6-35B-A3B:peft:262144":{"id":"Qwen/Qwen3.6-35B-A3B:peft:262144","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":1.335,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.5-35B-A3B-Base":{"id":"Qwen/Qwen3.5-35B-A3B-Base","name":"Qwen3.5 35B-A3B Base","description":"Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.54,"output":1.335,"cache_read":0.108},"canonical_model_id":"alibaba/qwen3.5-35b-a3b-base"},"Qwen/Qwen3.8-27B:peft:262144":{"id":"Qwen/Qwen3.8-27B:peft:262144","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6975,"output":5.595,"cache_read":0.1395},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.5-9B-Base":{"id":"Qwen/Qwen3.5-9B-Base","name":"Qwen3.5 9B Base","description":"Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.3,"output":0.807,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3.5-9b-base"},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen3 8B","description":"Dense 8B Qwen3 model with switchable thinking for chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.195,"output":0.6,"cache_read":0.039},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.45,"output":1.335,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.3,"output":0.807,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.6975,"output":5.595,"cache_read":0.1395},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.65,"output":7.5,"cache_read":0.33},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"thinkingmachines/Inkling-Small:peft:262144":{"id":"thinkingmachines/Inkling-Small:peft:262144","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.58,"output":1.44,"cache_read":0.116},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/Inkling:peft:262144:sampling-nvfp4":{"id":"thinkingmachines/Inkling:peft:262144:sampling-nvfp4","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling-Small:peft:262144:sampling-nvfp4":{"id":"thinkingmachines/Inkling-Small:peft:262144:sampling-nvfp4","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling:peft:262144":{"id":"thinkingmachines/Inkling:peft:262144","name":"Inkling (256K)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.58,"output":1.44,"cache_read":0.116},"canonical_model_id":"thinkingmachines/inkling-small"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.18,"output":0.45,"cache_read":0.036},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.33,"output":0.84,"cache_read":0.066},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-120b:peft:131072":{"id":"openai/gpt-oss-120b:peft:131072","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.33,"output":0.84,"cache_read":0.066},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":1.695,"output":4.215,"cache_read":0.339},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V4.1-Flash:peft:262144":{"id":"deepseek-ai/DeepSeek-V4.1-Flash:peft:262144","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.6,"cache_read":0.12},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4.1-Flash-1017:peft:262144":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017:peft:262144","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.6,"cache_read":0.12},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":2.205,"output":5.49,"cache_read":0.441},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.6:peft:131072":{"id":"moonshotai/Kimi-K2.6:peft:131072","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":2.205,"output":5.49,"cache_read":0.441},"canonical_model_id":"moonshotai/kimi-k2.6"},"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16:peft:262144","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.195,"output":0.495,"cache_read":0.039},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16:peft:262144","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16:peft:262144","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.184,"output":0.495,"cache_read":0.0368},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.8,"output":6.225,"cache_read":0.36},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.184,"output":0.495,"cache_read":0.0368},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16:peft:262144","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.8,"output":6.225,"cache_read":0.36},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"zai-org/GLM-5.3:peft:262144":{"id":"zai-org/GLM-5.3:peft:262144","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":3.78,"output":12.15,"cache_read":0.756},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash:peft:262144":{"id":"zai-org/GLM-5.3-Flash:peft:262144","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.84,"output":2.11,"cache_read":0.168},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"lilac":{"id":"lilac","env":["LILAC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.getlilac.com/v1","name":"Lilac","doc":"https://docs.getlilac.com/inference/models","models":{"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":262100},"status":"deprecated","cost":{"input":0.11,"output":0.35},"canonical_model_id":"google/gemma-4-31b-it"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.7,"output":3.5,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimaxai/minimax-m3":{"id":"minimaxai/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"status":"deprecated","cost":{"input":0.28,"output":1.1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":524288},"status":"deprecated","cost":{"input":0.9,"output":3,"cache_read":0.27},"canonical_model_id":"zhipuai/glm-5.2"}}},"ofox":{"id":"ofox","env":["OFOX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ofox.ai/v1","name":"Ofox","doc":"https://ofox.ai/docs","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash-0423":{"id":"deepseek/deepseek-v4-flash-0423","name":"DeepSeek V4 Flash 0423","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.19,"output":0.51,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-v4-pro-0423":{"id":"deepseek/deepseek-v4-pro-0423","name":"DeepSeek V4 Pro 0423","description":"DeepSeek V4 Pro initial snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v4-pro-0423"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.29,"output":0.43,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v3.2"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.1-fast":{"id":"x-ai/grok-4.1-fast","name":"Grok 4.1 Fast","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":4,"output":12,"cache_read":0.4},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.5-lightning":{"id":"minimax/minimax-m2.5-lightning","name":"MiniMax-M2.5 Lightning","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/m2-her":{"id":"minimax/m2-her","name":"MiniMax-M2 Her","description":"MiniMax M2 variant tuned for conversational and character-driven agent interactions","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":2048},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2-Her"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.1-lightning":{"id":"minimax/minimax-m2.1-lightning","name":"MiniMax-M2.1 Lightning","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"volcengine/doubao-seed-2.0-pro":{"id":"volcengine/doubao-seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.67,"output":3.36,"cache_read":0.14,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"volcengine/doubao-seed-2.0-lite":{"id":"volcengine/doubao-seed-2.0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.13,"output":0.76,"cache_read":0.03,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"volcengine/doubao-seed-2.0-code":{"id":"volcengine/doubao-seed-2.0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.67,"output":3.36,"cache_read":0.14,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"volcengine/doubao-seed-2.1-turbo":{"id":"volcengine/doubao-seed-2.1-turbo","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.3536,"output":1.7696,"cache_read":0.068,"cache_write":0.0019},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"volcengine/doubao-seed-2.1-pro":{"id":"volcengine/doubao-seed-2.1-pro","name":"Seed 2.1 Pro","description":"Flagship ByteDance Seed 2.1 model for complex multimodal reasoning, coding, and agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.7072,"output":3.536,"cache_read":0.1416,"cache_write":0.002},"canonical_model_id":"bytedance-seed/seed-2.1-pro"},"volcengine/doubao-seed-character":{"id":"volcengine/doubao-seed-character","name":"Seed Character","description":"ByteDance Seed model optimized for character-driven dialogue and consistent conversational behavior","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.177,"output":0.884,"cache_read":0.024,"cache_write":0.0025},"canonical_model_id":"bytedance-seed/seed-character"},"volcengine/doubao-seed-2.0-mini":{"id":"volcengine/doubao-seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.06,"output":0.56,"cache_read":0.02,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"volcengine/doubao-seed-evolving":{"id":"volcengine/doubao-seed-evolving","name":"Seed Evolving","description":"Rolling ByteDance Seed model for rapidly updated reasoning, coding, and agent capabilities","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.884,"output":4.42,"cache_read":0.177,"cache_write":0.0025},"canonical_model_id":"bytedance-seed/seed-evolving"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":4.5,"input_audio":1.25},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.0415,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":4.5,"input_audio":2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.0415,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":1,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":1,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.0415,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.083,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":1,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.05,"output":0.4,"cache_read":0.01},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"cache_write":6.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5,"cache_write":6.25}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1131072,"output":131072},"cost":{"input":0.5,"output":1.71,"cache_read":0.043,"cache_write":0.63},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen-vl-max":{"id":"qwen/qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8000},"cost":{"input":0.23,"output":0.58,"cache_read":0.023},"canonical_model_id":"alibaba/qwen-vl-max"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":2.5,"cache_read":0.06,"cache_write":0.27},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1064000,"output":64000},"cost":{"input":1.71,"output":5.14,"cache_read":0.17,"cache_write":2.14},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":1.83,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.8,"output":9,"cache_read":0.2,"cache_write":1},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":0.36,"output":1.43,"cache_read":0.072},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen-flash":{"id":"qwen/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.022,"output":0.22,"cache_read":0.0043,"cache_write":0.027},"canonical_model_id":"alibaba/qwen-flash"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":2.29,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.6-max-preview":{"id":"qwen/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":2.15,"output":12.86,"cache_read":0.2,"cache_write":1.17},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"qwen/qwen-turbo":{"id":"qwen/qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.043,"output":0.09,"cache_read":0.0086},"canonical_model_id":"alibaba/qwen-turbo"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.55,"output":3.5,"cache_read":0.55},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.71,"output":5.14,"cache_read":0.17,"cache_write":2.14},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.71,"output":5.14,"cache_read":0.17,"cache_write":2.14},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.43,"output":2.57},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11,"output":0.39,"cache_read":0.011,"cache_write":0.14},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.12,"output":0.29,"cache_read":0.023},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1064000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen-max":{"id":"qwen/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8000},"cost":{"input":0.35,"output":1.38,"cache_read":0.069},"canonical_model_id":"alibaba/qwen-max"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":2.05,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.31},"canonical_model_id":"alibaba/qwen3.6-flash"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.7-flashx":{"id":"z-ai/glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.072,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.4,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"bailian/qwen3.8-27b":{"id":"bailian/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1131072,"output":131072},"cost":{"input":0.45,"output":3.2,"cache_read":0.05,"cache_write":0.5625},"canonical_model_id":"alibaba/qwen3.8-27b"},"bailian/qwen-vl-max":{"id":"bailian/qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.23,"output":0.58,"cache_read":0.046},"canonical_model_id":"alibaba/qwen-vl-max"},"bailian/qwen3.6-plus":{"id":"bailian/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"bailian/qwen3-coder-flash":{"id":"bailian/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":2.5,"cache_read":0.06,"cache_write":0.27},"canonical_model_id":"alibaba/qwen3-coder-flash"},"bailian/qwen3.7-max":{"id":"bailian/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"bailian/qwen3.5-35b-a3b":{"id":"bailian/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.29,"output":1.83,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"bailian/qwen3-coder-plus":{"id":"bailian/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.8,"output":9,"cache_read":0.2,"cache_write":1},"canonical_model_id":"alibaba/qwen3-coder-plus"},"bailian/qwen3-max":{"id":"bailian/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.36,"output":1.43,"cache_read":0.072},"canonical_model_id":"alibaba/qwen3-max"},"bailian/qwen-flash":{"id":"bailian/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.022,"output":0.22,"cache_read":0.0043,"cache_write":0.027},"canonical_model_id":"alibaba/qwen-flash"},"bailian/qwen3.5-122b-a10b":{"id":"bailian/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":2.29,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"bailian/qwen3.6-max-preview":{"id":"bailian/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":2.15,"output":12.86,"cache_read":0.2,"cache_write":1.17},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"bailian/qwen-turbo":{"id":"bailian/qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.05,"output":0.09,"cache_read":0.0086},"canonical_model_id":"alibaba/qwen-turbo"},"bailian/qwen3.5-397b-a17b":{"id":"bailian/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.55,"output":3.5,"cache_read":0.55},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"bailian/qwen3.8-max":{"id":"bailian/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"bailian/qwen3.8-max-0902":{"id":"bailian/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"bailian/qwen3.5-plus":{"id":"bailian/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"bailian/qwen3.6-27b":{"id":"bailian/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.6-27b"},"bailian/qwen3.8-flash":{"id":"bailian/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"bailian/qwen-plus":{"id":"bailian/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.12,"output":0.29,"cache_read":0.023},"canonical_model_id":"alibaba/qwen-plus"},"bailian/qwen3.7-plus":{"id":"bailian/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"bailian/qwen-max":{"id":"bailian/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.35,"output":1.38,"cache_read":0.069},"canonical_model_id":"alibaba/qwen-max"},"bailian/qwen3.5-flash":{"id":"bailian/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"bailian/qwen3.5-27b":{"id":"bailian/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.29,"output":2.05,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-27b"},"bailian/qwen3-coder-next":{"id":"bailian/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"bailian/qwen3.6-flash":{"id":"bailian/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.31},"canonical_model_id":"alibaba/qwen3.6-flash"}}},"qiniu-ai":{"id":"qiniu-ai","env":["QINIU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.qnaigc.com/v1","name":"Qiniu","doc":"https://developer.qiniu.com/aitokenapi","models":{"qwen2.5-vl-72b-instruct":{"id":"qwen2.5-vl-72b-instruct","name":"Qwen 2.5 VL 72B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":2.31884058,"output":6.95652174},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"doubao-1.5-pro-32k":{"id":"doubao-1.5-pro-32k","name":"Doubao 1.5 Pro 32k","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":12000},"status":"deprecated","cost":{"input":0.11594203,"output":0.28985507}},"doubao-seed-1.6-thinking":{"id":"doubao-seed-1.6-thinking","name":"Doubao-Seed 1.6 Thinking","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-15","last_updated":"2025-08-15","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.11594203,"output":1.15942029,"tiers":[{"input":0.17391304,"output":2.31884058,"tier":{"type":"context","size":32000}},{"input":0.34782609,"output":3.47826087,"tier":{"type":"context","size":128000}}]}},"doubao-seed-2.0-pro":{"id":"doubao-seed-2.0-pro","name":"Doubao Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.46376812,"output":2.31884058,"cache_read":0.09275362,"tiers":[{"input":0.69565217,"output":3.47826087,"cache_read":0.13913043,"tier":{"type":"context","size":32000}},{"input":1.39130435,"output":6.95652174,"cache_read":0.27826087,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"qwen-vl-max-2025-01-25":{"id":"qwen-vl-max-2025-01-25","name":"Qwen VL-MAX-2025-01-25","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.43478261,"output":1.30434783}},"qwen3-vl-30b-a3b-thinking":{"id":"qwen3-vl-30b-a3b-thinking","name":"Qwen3-Vl 30b A3b Thinking","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-09","last_updated":"2026-02-09","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.10869565,"output":1.08695652}},"gemini-3.0-flash-preview":{"id":"gemini-3.0-flash-preview","name":"Gemini 3.0 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-3-flash-preview"},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen 3 235B A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":4096},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"alibaba/qwen3-32b"},"kimi-k2":{"id":"kimi-k2","name":"Kimi K2","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"moonshotai/kimi-k2"},"gemini-3.0-pro-image-preview":{"id":"gemini-3.0-pro-image-preview","type":"image","name":"Gemini 3.0 Pro Image Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-3-pro-image-preview"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-2.5-pro"},"doubao-1.5-thinking-pro":{"id":"doubao-1.5-thinking-pro","name":"Doubao 1.5 Thinking Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058}},"claude-4.0-sonnet":{"id":"claude-4.0-sonnet","name":"Claude 4.0 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","canonical_model_id":"anthropic/claude-sonnet-4-0"},"qwen3-next-80b-a3b-thinking":{"id":"qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B Thinking","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.14492754,"output":1.44927536},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"claude-4.1-opus":{"id":"claude-4.1-opus","name":"Claude 4.1 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","canonical_model_id":"anthropic/claude-opus-4-1"},"kling-v2-6":{"id":"kling-v2-6","type":"video","name":"Kling-V2 6","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-13","last_updated":"2026-01-13","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":99999999,"output":99999999}},"doubao-seed-2.0-lite":{"id":"doubao-seed-2.0-lite","name":"Doubao Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.08695652,"output":0.52173913,"cache_read":0.0173913,"tiers":[{"input":0.13043478,"output":0.7826087,"cache_read":0.02608696,"tier":{"type":"context","size":32000}},{"input":0.26086957,"output":1.56521739,"cache_read":0.05217391,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","canonical_model_id":"openai/gpt-oss-20b"},"gemini-3.0-pro-preview":{"id":"gemini-3.0-pro-preview","name":"Gemini 3.0 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-3-pro-preview"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235b A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen-max-2025-01-25":{"id":"qwen-max-2025-01-25","name":"Qwen2.5-Max-2025-01-25","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.34782609,"output":1.39130435},"canonical_model_id":"alibaba/qwen-max"},"claude-4.5-haiku":{"id":"claude-4.5-haiku","name":"Claude 4.5 Haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","canonical_model_id":"anthropic/claude-haiku-4-5"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.86956522,"output":3.47826087,"tiers":[{"input":1.44927536,"output":5.79710145,"tier":{"type":"context","size":32000}},{"input":2.17391304,"output":8.69565217,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"qwen3-235b-a22b-thinking-2507":{"id":"qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22B Thinking 2507","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"status":"deprecated","cost":{"input":0.28985507,"output":2.89855072},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"doubao-seed-2.0-code":{"id":"doubao-seed-2.0-code","name":"Doubao Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.46376812,"output":2.31884058,"cache_read":0.09275362,"tiers":[{"input":0.69565217,"output":3.47826087,"cache_read":0.13913043,"tier":{"type":"context","size":32000}},{"input":1.39130435,"output":6.95652174,"cache_read":0.27826087,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"claude-3.5-haiku":{"id":"claude-3.5-haiku","name":"Claude 3.5 Haiku","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07-31","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"glm-4.5":{"id":"glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"zhipuai/glm-4.5"},"qwen-turbo":{"id":"qwen-turbo","name":"Qwen-Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":4096},"status":"deprecated","cost":{"input":0.04347826,"output":0.08695652},"canonical_model_id":"alibaba/qwen-turbo"},"doubao-seed-1.6-flash":{"id":"doubao-seed-1.6-flash","name":"Doubao-Seed 1.6 Flash","description":"Low-latency ByteDance Seed model for high-throughput chat, extraction, and lightweight tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.02173913,"output":0.2173913,"tiers":[{"input":0.04347826,"output":0.43478261,"tier":{"type":"context","size":32000}},{"input":0.08695652,"output":0.86956522,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"qwen3-30b-a3b-instruct-2507":{"id":"qwen3-30b-a3b-instruct-2507","name":"Qwen3 30b A3b Instruct 2507","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.10869565,"output":0.43478261},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.17391304,"output":1.04347826,"tiers":[{"input":0.43478261,"output":2.60869565,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-v3":{"id":"deepseek-v3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"deepseek/deepseek-v3"},"qwen3-max-preview":{"id":"qwen3-max-preview","name":"Qwen3 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-05","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.86956522,"output":3.47826087,"tiers":[{"input":1.44927536,"output":5.79710145,"tier":{"type":"context","size":32000}},{"input":2.17391304,"output":8.69565217,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max-preview"},"deepseek-r1-0528":{"id":"deepseek-r1-0528","name":"DeepSeek-R1-0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"deepseek/deepseek-r1-0528"},"claude-3.7-sonnet":{"id":"claude-3.7-sonnet","name":"Claude 3.7 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"status":"deprecated","canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"claude-4.0-opus":{"id":"claude-4.0-opus","name":"Claude 4.0 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","canonical_model_id":"anthropic/claude-opus-4-0"},"deepseek-v3.1":{"id":"deepseek-v3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":1.73913043},"canonical_model_id":"deepseek/deepseek-v3.1"},"doubao-seed-1.6":{"id":"doubao-seed-1.6","name":"Doubao-Seed 1.6","description":"ByteDance Seed model for long-context reasoning, instruction following, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.11594203,"output":1.15942029,"tiers":[{"input":0.17391304,"output":2.31884058,"tier":{"type":"context","size":32000}},{"input":0.34782609,"output":3.47826087,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash"},"claude-4.5-sonnet":{"id":"claude-4.5-sonnet","name":"Claude 4.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","canonical_model_id":"anthropic/claude-sonnet-4-5"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":4096},"status":"deprecated","cost":{"input":0.10869565,"output":0.43478261},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"gemini-2.0-flash-lite":{"id":"gemini-2.0-flash-lite","name":"Gemini 2.0 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-2.0-flash-lite"},"doubao-1.5-vision-pro":{"id":"doubao-1.5-vision-pro","name":"Doubao 1.5 Vision Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"status":"deprecated","cost":{"input":0.43478261,"output":1.30434783}},"doubao-seed-2.0-mini":{"id":"doubao-seed-2.0-mini","name":"Doubao Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.02898551,"output":0.28985507,"cache_read":0.0057971,"tiers":[{"input":0.05797101,"output":0.57971014,"cache_read":0.0115942,"tier":{"type":"context","size":32000}},{"input":0.11594203,"output":1.15942029,"cache_read":0.02318841,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"MiniMax-M1":{"id":"MiniMax-M1","name":"MiniMax M1","description":"Open hybrid-attention MoE reasoning model with a million-token context and an 80K thinking budget","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":80000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"minimax/minimax-m1-80k"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":4096},"status":"deprecated","canonical_model_id":"zhipuai/glm-4.5-air"},"claude-3.5-sonnet":{"id":"claude-3.5-sonnet","name":"Claude 3.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","canonical_model_id":"anthropic/claude-3-5-sonnet-20241022"},"mimo-v2-flash":{"id":"mimo-v2-flash","name":"Mimo-V2-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"xiaomi/mimo-v2-flash"},"qwen3-30b-a3b-thinking-2507":{"id":"qwen3-30b-a3b-thinking-2507","name":"Qwen3 30b A3b Thinking 2507","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":126000,"output":32000},"status":"deprecated","cost":{"input":0.10869565,"output":1.08695652}},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":4096},"status":"deprecated","cost":{"input":0.86956522,"output":3.47826087,"tiers":[{"input":1.30434783,"output":5.2173913,"tier":{"type":"context","size":32000}},{"input":2.17391304,"output":8.69565217,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.14492754,"output":0.57971014},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"deepseek-r1":{"id":"deepseek-r1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"deepseek/deepseek-r1"},"claude-4.5-opus":{"id":"claude-4.5-opus","name":"Claude 4.5 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":200000},"status":"deprecated","canonical_model_id":"anthropic/claude-opus-4-5"},"gemini-2.0-flash":{"id":"gemini-2.0-flash","name":"Gemini 2.0 Flash","description":"Earlier Gemini Flash workhorse for responsive multimodal apps and tool use","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-2.0-flash"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Gemini 2.5 Flash Image","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash-image"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash-lite"},"deepseek/deepseek-v3.1-terminus":{"id":"deepseek/deepseek-v3.1-terminus","name":"DeepSeek/DeepSeek-V3.1-Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":1.73913043}},"deepseek/deepseek-v3.1-terminus-thinking":{"id":"deepseek/deepseek-v3.1-terminus-thinking","name":"DeepSeek/DeepSeek-V3.1-Terminus-Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":1.73913043}},"deepseek/deepseek-v3.2-251201":{"id":"deepseek/deepseek-v3.2-251201","name":"Deepseek/DeepSeek-V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":0.43478261,"cache_read":0.02898551},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v3.2-exp-thinking":{"id":"deepseek/deepseek-v3.2-exp-thinking","name":"DeepSeek/DeepSeek-V3.2-Exp-Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":0.43478261}},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek/DeepSeek-V3.2-Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":0.43478261}},"deepseek/deepseek-math-v2":{"id":"deepseek/deepseek-math-v2","name":"Deepseek/Deepseek-Math-V2","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-04","last_updated":"2025-12-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":160000,"output":160000},"status":"deprecated"},"x-ai/grok-4.1-fast-non-reasoning":{"id":"x-ai/grok-4.1-fast-non-reasoning","name":"X-Ai/Grok 4.1 Fast Non Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast"},"x-ai/grok-code-fast-1":{"id":"x-ai/grok-code-fast-1","name":"x-AI/Grok-Code-Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":10000},"status":"deprecated","canonical_model_id":"xai/grok-code-fast-1"},"x-ai/grok-4-fast":{"id":"x-ai/grok-4-fast","name":"x-AI/Grok-4-Fast","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-20","last_updated":"2025-09-20","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated"},"x-ai/grok-4.1-fast":{"id":"x-ai/grok-4.1-fast","name":"x-AI/Grok-4.1-Fast","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated"},"x-ai/grok-4-fast-reasoning":{"id":"x-ai/grok-4-fast-reasoning","name":"X-Ai/Grok-4-Fast-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4-fast-reasoning"},"x-ai/grok-4.1-fast-reasoning":{"id":"x-ai/grok-4.1-fast-reasoning","name":"X-Ai/Grok 4.1 Fast Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":20000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast-reasoning"},"x-ai/grok-4-fast-non-reasoning":{"id":"x-ai/grok-4-fast-non-reasoning","name":"X-Ai/Grok-4-Fast-Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4-fast-non-reasoning"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"Minimax/Minimax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"cost":{"input":0.30434783,"output":1.2173913},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"Minimax/Minimax-M2.5 Highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"status":"deprecated","cost":{"input":0.60869565,"output":2.43478261},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"Minimax/Minimax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"status":"deprecated","cost":{"input":0.30434783,"output":1.2173913},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"Minimax/Minimax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":0.30434783,"output":1.2173913},"canonical_model_id":"minimax/MiniMax-M2"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Stepfun/Step-3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"input":256000,"output":4096},"status":"deprecated","canonical_model_id":"stepfun/step-3.5-flash"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"OpenAI/GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5":{"id":"openai/gpt-5","name":"OpenAI/GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","canonical_model_id":"openai/gpt-5"},"stepfun-ai/gelab-zero-4b-preview":{"id":"stepfun-ai/gelab-zero-4b-preview","name":"Stepfun-Ai/Gelab Zero 4b Preview","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":4096},"status":"deprecated"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Moonshotai/Kimi-K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0.57971014,"output":3.04347826},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":100000},"status":"deprecated","canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":100000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"Z-Ai/GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.57971014,"output":2.60869565,"tiers":[{"input":0.86956522,"output":3.1884058,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"Z-Ai/GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":200000},"status":"deprecated","cost":{"input":0.45913043,"output":1.81565217},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"Z-AI/GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":200000},"status":"deprecated","cost":{"input":1.04347826,"output":1.82608696},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/autoglm-phone-9b":{"id":"z-ai/autoglm-phone-9b","name":"Z-Ai/Autoglm Phone 9b","description":"GLM vision model for visual reasoning, documents, and multimodal agents","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":12800,"output":4096},"status":"deprecated"},"meituan/longcat-flash-lite":{"id":"meituan/longcat-flash-lite","name":"Meituan/Longcat-Flash-Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-06","last_updated":"2026-02-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":320000},"status":"deprecated","cost":{"input":0,"output":0}},"meituan/longcat-flash-chat":{"id":"meituan/longcat-flash-chat","name":"Meituan/Longcat-Flash-Chat","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-11-05","last_updated":"2025-11-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"status":"deprecated"}}},"standardcompute":{"id":"standardcompute","env":["STANDARDCOMPUTE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stdcmpt.com/v1","name":"Standard Compute","doc":"https://docs.standardcompute.com/models","models":{"standardcompute":{"id":"standardcompute","name":"Standard Compute","description":"Flat-rate smart-routing gateway: one model id, each request routed across the customer's enabled models from a curated ~1M-context catalog (Claude, GPT, Gemini, DeepSeek, GLM, Qwen, Kimi, MiniMax, Grok, MiMo) or pinned to a user-selected model","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-03-01","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"}},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.14,"output":0.42,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.044,"output":2.088,"cache_read":0.2088},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.0826,"output":0.1652,"cache_read":0.01652},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.122,"output":3.366,"cache_read":0.1122},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":30000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":0.375,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.8-flash"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"mistralai/mistral-large-4-0":{"id":"mistralai/mistral-large-4-0","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.425,"output":2.55,"cache_read":0.085,"cache_write":0.53125},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.32,"output":1.28,"cache_read":0.064,"cache_write":0.4,"tiers":[{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.5625,"output":1.8,"cache_read":0.105},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.08,"output":0.5,"cache_read":0.012},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"hetzner":{"id":"hetzner","env":["HETZNER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.hetzner.com/api/v1","name":"Hetzner","doc":"https://experiments.hetzner.com/docs/inference","models":{"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8-27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"beta","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.6-35B-A3B-FP8":{"id":"Qwen/Qwen3.6-35B-A3B-FP8","name":"Qwen3.6 35B A3B FP8","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"beta","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"llmgateway":{"id":"llmgateway","env":["LLMGATEWAY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.llmgateway.io/v1","name":"DevPass (LLM Gateway)","doc":"https://llmgateway.io/docs","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"cost":{"input":0.573,"output":2.58,"cache_read":0.144,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"qwen3-vl-235b-a22b-thinking":{"id":"qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"codestral-2508":{"id":"codestral-2508","name":"Codestral","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16384},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-2508"},"gpt-live-transcribe":{"id":"gpt-live-transcribe","type":"transcription","name":"GPT-Live-Transcribe","description":"Low-latency streaming speech-to-text model with keyword and language hints","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-live-transcribe"},"ernie-4.5-vl-424b-a47b":{"id":"ernie-4.5-vl-424b-a47b","name":"ERNIE 4.5 VL 424B A47B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":123000,"output":123000},"cost":{"input":0.42,"output":1.25}},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.1-flash-lite"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":384000},"cost":{"input":0.135,"output":0.54,"cache_read":0.0027},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"seed-1-6-flash-250715":{"id":"seed-1-6-flash-250715","name":"Seed 1.6 Flash (250715)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.07,"output":0.3,"cache_read":0.015}},"glm-5-turbo":{"id":"glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.5,"output":2,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-5.3"},"seed-1-8-251228":{"id":"seed-1-8-251228","name":"Seed 1.8 (251228)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.135,"output":0.4},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"llama-3.2-11b-instruct":{"id":"llama-3.2-11b-instruct","name":"Llama 3.2 11B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.07,"output":0.33}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"gpt-5.2-pro":{"id":"gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":228700,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"qwen3-vl-235b-a22b-instruct":{"id":"qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":384000},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro"},"grok-4-20-beta-0309-non-reasoning":{"id":"grok-4-20-beta-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":0.07,"output":0.27},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"reka-edge-2603":{"id":"reka-edge-2603","name":"Reka Edge 2603","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-09","last_updated":"2026-10-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":16384},"cost":{"input":0.1,"output":0.1}},"seed-2-0-mini-260428":{"id":"seed-2-0-mini-260428","name":"Seed 2.0 Mini (260428)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.4,"cache_read":0.02}},"fugu-max":{"id":"fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"gemini-pro-latest":{"id":"gemini-pro-latest","name":"Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-27","last_updated":"2026-02-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2}},"sonar":{"id":"sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":130000,"output":4096},"cost":{"input":0.25,"output":2.5,"cache_write":0.25},"canonical_model_id":"perplexity/sonar"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.36,"output":0.87,"reasoning":8.4},"canonical_model_id":"alibaba/qwen3-32b"},"kimi-k2":{"id":"kimi-k2","name":"Kimi K2","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.57,"output":2.3,"cache_read":0.5}},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":1000000},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.0375},"canonical_model_id":"alibaba/qwen3.7-flash"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"qwen-plus-latest":{"id":"qwen-plus-latest","name":"Qwen Plus Latest","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-01-25","last_updated":"2025-01-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":8192},"cost":{"input":0.4,"output":1.2,"cache_read":0.08,"cache_write":0.5}},"qwen3-coder-flash":{"id":"qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":997952,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":3.75,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.574,"output":3.011,"cache_read":0.225},"canonical_model_id":"moonshotai/kimi-k2.5"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.12,"output":0.25,"cache_read":0.0025},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"Ministral 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"mistral/ministral-14b-2512"},"devstral-2512":{"id":"devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-2512"},"fugu-ultra-v2.0":{"id":"fugu-ultra-v2.0","name":"Fugu Ultra v2.0","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5}},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.04,"output":0.19,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-20b"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.06,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"auto":{"id":"auto","name":"Auto Route","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0,"output":0}},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct (2507)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.09,"output":0.58},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"seed-2-0-pro-260328":{"id":"seed-2-0-pro-260328","name":"Seed 2.0 Pro (260328)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"glm-5.2-fast":{"id":"glm-5.2-fast","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.156,"output":6.776,"cache_read":0.429},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":1.6,"reasoning":4.8,"cache_read":0.04,"cache_write":0.25},"canonical_model_id":"alibaba/qwen3-vl-plus"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.6-flash"},"granite-4.2-8b":{"id":"granite-4.2-8b","name":"Granite 4.2 8B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"ministral-3b-2512":{"id":"ministral-3b-2512","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"mistral/ministral-3b-2512"},"glm-4.7-flashx":{"id":"glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.737,"output":3.08,"cache_read":0.143},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.08,"output":0.32,"cache_read":0.017,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax-m2.5-highspeed":{"id":"minimax-m2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.845,"output":3.38,"cache_read":0.6,"cache_write":3.75},"canonical_model_id":"alibaba/qwen3-max"},"qwen-flash":{"id":"qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.05,"output":0.4,"cache_read":0.01,"cache_write":0.0625},"canonical_model_id":"alibaba/qwen-flash"},"custom":{"id":"custom","name":"Custom Model","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0,"output":0}},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen3-235b-a22b-thinking-2507":{"id":"qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22B Thinking (2507)","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":8192},"cost":{"input":0.3,"output":3}},"nemotron-3-ultra-550b":{"id":"nemotron-3-ultra-550b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":0.5,"output":2.2,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"mistral-medium-3-5":{"id":"mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":384000},"cost":{"input":0.076,"output":0.153,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"seed-2-0-lite-260428":{"id":"seed-2-0-lite-260428","name":"Seed 2.0 Lite (260428)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"gpt-transcribe":{"id":"gpt-transcribe","type":"transcription","name":"GPT-Transcribe","description":"High-accuracy speech-to-text model for audio files and committed Realtime turns","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-transcribe"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.45,"output":2,"cache_read":0.19,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"qwen3.5-122b-a10b":{"id":"qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":1050000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"minimax-m2.1":{"id":"minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.27,"output":1.1},"canonical_model_id":"minimax/MiniMax-M2.1"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"llama-3.2-3b-instruct":{"id":"llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-09-18","last_updated":"2024-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.03,"output":0.05}},"glm-4-32b-0414-128k":{"id":"glm-4-32b-0414-128k","name":"GLM-4 32B (0414-128k)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.1}},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"ling-3.0-flash-vl":{"id":"ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":1050000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"glm-5.3-fast":{"id":"glm-5.3-fast","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.56},"canonical_model_id":"zhipuai/glm-5.3"},"minimax-m2.1-lightning":{"id":"minimax-m2.1-lightning","name":"MiniMax M2.1 Lightning","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.12,"output":0.48}},"glm-4.6v-flashx":{"id":"glm-4.6v-flashx","name":"GLM-4.6V FlashX","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.04,"output":0.4,"cache_read":0.004}},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"gpt-4o-mini-transcribe":{"id":"gpt-4o-mini-transcribe","name":"GPT-4o Mini Transcribe","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":1.25,"output":5}},"hy-mt2-plus":{"id":"hy-mt2-plus","name":"Hy-MT2 Plus","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"Hy","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.074,"output":0.295}},"qwen3-235b-a22b-fp8":{"id":"qwen3-235b-a22b-fp8","name":"Qwen3 235B A22B FP8","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":8192},"cost":{"input":0.2,"output":0.8}},"glm-4.5-airx":{"id":"glm-4.5-airx","name":"GLM-4.5 AirX","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.1,"output":4.5,"cache_read":0.22}},"minimax-text-01":{"id":"minimax-text-01","name":"MiniMax Text 01","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2,"output":1.1}},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.54,"output":4.631,"cache_read":0.198,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"grok-4-7":{"id":"grok-4-7","name":"Grok 4.7","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"llama-3-70b-instruct":{"id":"llama-3-70b-instruct","name":"Llama 3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-04-18","last_updated":"2024-04-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8000},"cost":{"input":0.51,"output":0.74}},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.05,"cache_read":0.13},"canonical_model_id":"moonshotai/kimi-k2.6"},"mistral-small-2506":{"id":"mistral-small-2506","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/mistral-small-2506"},"fugu-ultra":{"id":"fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"ling-3.0-flash":{"id":"ling-3.0-flash","name":"InclusionAI Ling 3.0 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"qwen3-14b":{"id":"qwen3-14b","name":"Qwen3 14B","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.076,"output":0.228}},"llama-3.1-70b-instruct":{"id":"llama-3.1-70b-instruct","name":"Llama 3.1 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"status":"beta","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"grok-4":{"id":"grok-4","name":"Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":3,"output":15,"cache_read":0.75}},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-4o-transcribe":{"id":"gpt-4o-transcribe","name":"GPT-4o Transcribe","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":2.5,"output":10}},"kimi-k3.1-pro-highspeed":{"id":"kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4}},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":65536},"cost":{"input":0.15,"output":1.35},"canonical_model_id":"alibaba/qwen3.6-27b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.605,"output":2.123,"cache_read":0.154,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"grok-4-3":{"id":"grok-4-3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"seed-1-6-250915":{"id":"seed-1-6-250915","name":"Seed 1.6 (250915)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"qwen-plus":{"id":"qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.4,"output":1.2,"reasoning":4,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen-plus"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.088,"output":0.25,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen35-397b-a17b":{"id":"qwen35-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.095,"output":0.285},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8 27B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":32768},"cost":{"input":0.08,"output":0.35,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen-max":{"id":"qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4},"canonical_model_id":"alibaba/qwen-max"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.032,"output":0.14,"cache_read":0.032},"canonical_model_id":"openai/gpt-oss-120b"},"grok-4-20-reasoning":{"id":"grok-4-20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"mistral-large-latest":{"id":"mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"cost":{"input":4,"output":12},"canonical_model_id":"mistral/mistral-large-latest"},"qwen3-vl-30b-a3b-instruct":{"id":"qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.25,"cache_read":0.01},"canonical_model_id":"google/gemma-4-31b-it"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"glm-4.5-x":{"id":"glm-4.5-x","name":"GLM-4.5 X","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"beta","cost":{"input":2.2,"output":8.9,"cache_read":0.45}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-2603"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"minimax-m2":{"id":"minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.2,"output":1,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2"},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.8-flash"},"grok-4-20-non-reasoning":{"id":"grok-4-20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"minimax-m2.7-highspeed":{"id":"minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"llama-4-scout-17b-instruct":{"id":"llama-4-scout-17b-instruct","name":"Llama 4 Scout 17B Instruct","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":2048},"cost":{"input":0.18,"output":0.59},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"grok-build-0-1":{"id":"grok-build-0-1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"grok-4-20-beta-0309-reasoning":{"id":"grok-4-20-beta-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"muse-spark-1.2-contributor":{"id":"muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"grok-4-5":{"id":"grok-4-5","name":"Grok 4.5","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"seed-2-0-code-preview-260328":{"id":"seed-2-0-code-preview-260328","name":"Seed 2.0 Code Preview (260328)","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.108,"output":0.675,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"kimi-k3-fast":{"id":"kimi-k3-fast","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1040384,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash"},"llama-4-maverick-17b-instruct":{"id":"llama-4-maverick-17b-instruct","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":2048},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"seed-1-6-250615":{"id":"seed-1-6-250615","name":"Seed 1.6 (250615)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-25","last_updated":"2025-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"ministral-8b-2512":{"id":"ministral-8b-2512","name":"Ministral 8B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.15,"output":0.15},"canonical_model_id":"mistral/ministral-8b-2512"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.05,"cache_write":0.3125},"canonical_model_id":"alibaba/qwen3.6-flash"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"abacus":{"id":"abacus","env":["ABACUS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://routellm.abacus.ai/v1","name":"Abacus","doc":"https://abacus.ai/help/api","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"llama-3.3-70b-versatile":{"id":"llama-3.3-70b-versatile","name":"Llama 3.3 70B Versatile","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-terra"},"mimo-v2-pro":{"id":"mimo-v2-pro","name":"MiMo-V2-Pro","description":"Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":3,"cache_read":0.2},"canonical_model_id":"xiaomi/mimo-v2-pro"},"grok-4-0709":{"id":"grok-4-0709","name":"Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16384},"status":"deprecated","cost":{"input":3,"output":15}},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"google/gemini-2.5-pro"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":32768},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"gpt-5.2-chat-latest":{"id":"gpt-5.2-chat-latest","name":"GPT-5.2 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"Grok Code Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":1.5},"canonical_model_id":"xai/grok-code-fast-1"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.5,"output":7.5},"canonical_model_id":"alibaba/qwen3.7-max"},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"status":"deprecated","cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.3-chat-latest":{"id":"gpt-5.3-chat-latest","name":"GPT-5.3 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"kimi-k2-turbo-preview":{"id":"kimi-k2-turbo-preview","name":"Kimi K2 Turbo Preview","description":"Fast Kimi model for responsive chat, coding help, and agent loops","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"status":"deprecated","cost":{"input":0.15,"output":8}},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"claude-sonnet-4-20250514":{"id":"claude-sonnet-4-20250514","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-4o-mini"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-codex-max"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-7"},"route-llm":{"id":"route-llm","name":"RouteLLM","description":"RouteLLM routes prompts to an appropriate Abacus-backed text-generation model","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-01-01","last_updated":"2026-07-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":2,"output":10,"cache_read":0.1}},"claude-opus-4-20250514":{"id":"claude-opus-4-20250514","name":"Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-5.1-chat-latest":{"id":"gpt-5.1-chat-latest","name":"GPT-5.1 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"claude-3-7-sonnet-20250219":{"id":"claude-3-7-sonnet-20250219","name":"Claude Sonnet 3.7","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-5"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-11-17","last_updated":"2025-11-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":0.5}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":10},"canonical_model_id":"anthropic/claude-sonnet-5"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4},"canonical_model_id":"openai/o4-mini"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"gpt-5.3-codex-xhigh":{"id":"gpt-5.3-codex-xhigh","name":"GPT-5.3 Codex XHigh","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-codex"},"o3-pro":{"id":"o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"qwen-2.5-coder-32b":{"id":"qwen-2.5-coder-32b","name":"Qwen 2.5 Coder 32B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.79,"output":0.79},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"gpt-4o-2024-11-20":{"id":"gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"Grok 4 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"gemini-3.1-flash-image-preview":{"id":"gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":32768},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30},"canonical_model_id":"google/gemini-2.5-flash-image"},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.55,"output":1.66},"canonical_model_id":"deepseek/deepseek-v3.1"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.32,"output":3.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"status":"deprecated","cost":{"input":0.13,"output":0.6},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen2.5-72B-Instruct":{"id":"Qwen/Qwen2.5-72B-Instruct","name":"Qwen 2.5 72B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.11,"output":0.38}},"Qwen/QwQ-32B":{"id":"Qwen/QwQ-32B","name":"QwQ 32B","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":0.4},"canonical_model_id":"alibaba/qwq-32b"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M3"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":3.74,"output":9.36,"cache_read":0.748},"canonical_model_id":"thinkingmachines/inkling"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.037,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"status":"deprecated","cost":{"input":1.74,"output":3.48,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.27,"output":0.4},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":3,"output":7},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-01","last_updated":"2025-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.27,"output":1}},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"id":"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"status":"deprecated","cost":{"input":0.14,"output":0.59},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo":{"id":"meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo","name":"Llama 3.1 405B Instruct Turbo","description":"Compact Llama instruction model for fast chat and local deployment","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":3.5,"output":3.5}},"meta-llama/Meta-Llama-3.1-8B-Instruct":{"id":"meta-llama/Meta-Llama-3.1-8B-Instruct","name":"Llama 3.1 8B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.02,"output":0.05},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/Meta-Llama-3.3-70B-Instruct":{"id":"meta-llama/Meta-Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-4.5":{"id":"zai-org/GLM-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":96000},"status":"deprecated","cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.5"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"privatemode-ai":{"id":"privatemode-ai","env":["PRIVATEMODE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://localhost:8080/v1","name":"Privatemode AI","doc":"https://docs.privatemode.ai/api/overview","models":{"deepseek-ocr-2":{"id":"deepseek-ocr-2","name":"DeepSeek OCR 2","description":"High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"status":"beta","cost":{"input":0.8897,"output":1.4675,"cache_read":0.0924},"canonical_model_id":"deepseek/deepseek-ocr-2"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.791,"output":6.6326,"cache_read":0.1733},"canonical_model_id":"zhipuai/glm-5.3"},"whisper-large-v3":{"id":"whisper-large-v3","type":"transcription","name":"Whisper large-v3","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.01618,"output":0},"canonical_model_id":"openai/whisper-large-v3"},"glm-flash-latest":{"id":"glm-flash-latest","name":"GLM Flash (latest)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2311,"output":0.7511,"cache_read":0.0578},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen3-embedding-4b":{"id":"qwen3-embedding-4b","type":"embedding","name":"Qwen3-Embedding 4B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-06","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":2560},"cost":{"input":0.1502,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-4b"},"glm-latest":{"id":"glm-latest","name":"GLM (latest)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.791,"output":6.6326,"cache_read":0.1733},"canonical_model_id":"zhipuai/glm-5.3"},"voxtral-mini-3b":{"id":"voxtral-mini-3b","type":"transcription","name":"Voxtral Mini 3B","description":"Speech-to-text model for audio transcription, translation, and audio understanding","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.00462,"output":0},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"status":"deprecated","cost":{"input":1.791,"output":8.9436,"cache_read":0.1733},"canonical_model_id":"moonshotai/kimi-k2.6"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2311,"output":0.7511,"cache_read":0.0578},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.2311,"output":0.7511,"cache_read":0.0462},"canonical_model_id":"openai/gpt-oss-120b"},"kimi-latest":{"id":"kimi-latest","name":"Kimi (latest)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"status":"deprecated","cost":{"input":1.791,"output":8.9436,"cache_read":0.1733},"canonical_model_id":"moonshotai/kimi-k2.6"}}},"xiaomi-token-plan-cn":{"id":"xiaomi-token-plan-cn","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan-cn.xiaomimimo.com/v1","name":"Xiaomi Token Plan (China)","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.5-tts":{"id":"mimo-v2.5-tts","type":"speech","name":"MiMo-V2.5-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5-tts-voiceclone":{"id":"mimo-v2.5-tts-voiceclone","type":"speech","name":"MiMo-V2.5-TTS-VoiceClone","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voiceclone"},"mimo-v2.5-tts-voicedesign":{"id":"mimo-v2.5-tts-voicedesign","type":"speech","name":"MiMo-V2.5-TTS-VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voicedesign"}}},"minimax":{"id":"minimax","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.io/anthropic/v1","name":"MiniMax (minimax.io)","doc":"https://platform.minimax.io/docs/guides/quickstart","models":{"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375}},"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375}},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"}}},"clarifai":{"id":"clarifai","env":["CLARIFAI_PAT"],"npm":"@ai-sdk/openai-compatible","api":"https://api.clarifai.com/v2/ext/openai/v1","name":"Clarifai","doc":"https://docs.clarifai.com/compute/inference/","models":{"clarifai/main/models/mm-poly-8b":{"id":"clarifai/main/models/mm-poly-8b","name":"MM Poly 8B","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"mm-poly","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-06","last_updated":"2026-02-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"cost":{"input":0.658,"output":1.11}},"arcee_ai/AFM/models/trinity-mini":{"id":"arcee_ai/AFM/models/trinity-mini","name":"Trinity Mini","description":"Reasoning-tuned 26B MoE model with 3B active parameters for agents, tools, and multi-step workloads","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-12-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.045,"output":0.15},"canonical_model_id":"arcee-ai/trinity-mini"},"openai/chat-completion/models/gpt-oss-20b":{"id":"openai/chat-completion/models/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.045,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"openai/chat-completion/models/gpt-oss-120b-high-throughput":{"id":"openai/chat-completion/models/gpt-oss-120b-high-throughput","name":"GPT OSS 120B High Throughput","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.09,"output":0.36},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/deepseek-ocr/models/DeepSeek-OCR":{"id":"deepseek-ai/deepseek-ocr/models/DeepSeek-OCR","name":"DeepSeek OCR","description":"OCR model for extracting structured text from documents and screenshots","family":"deepseek","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2026-02-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.2,"output":0.7}},"moonshotai/chat-completion/models/Kimi-K2_6":{"id":"moonshotai/chat-completion/models/Kimi-K2_6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"mistralai/completion/models/Ministral-3-14B-Reasoning-2512":{"id":"mistralai/completion/models/Ministral-3-14B-Reasoning-2512","name":"Ministral 3 14B Reasoning 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-01","last_updated":"2025-12-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2.5,"output":1.7}},"mistralai/completion/models/Ministral-3-3B-Reasoning-2512":{"id":"mistralai/completion/models/Ministral-3-3B-Reasoning-2512","name":"Ministral 3 3B Reasoning 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12","last_updated":"2026-02-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.039,"output":0.54825}},"qwen/qwenCoder/models/Qwen3-Coder-30B-A3B-Instruct":{"id":"qwen/qwenCoder/models/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.11458,"output":0.74812},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"qwen/qwenLM/models/Qwen3-30B-A3B-Instruct-2507":{"id":"qwen/qwenLM/models/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwenLM/models/Qwen3-30B-A3B-Thinking-2507":{"id":"qwen/qwenLM/models/Qwen3-30B-A3B-Thinking-2507","name":"Qwen3 30B A3B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-31","last_updated":"2026-02-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.36,"output":1.3}},"minimaxai/chat-completion/models/MiniMax-M2_5-high-throughput":{"id":"minimaxai/chat-completion/models/MiniMax-M2_5-high-throughput","name":"MiniMax-M2.5 High Throughput","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"iteracompute":{"id":"iteracompute","env":["ITERACOMPUTE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.iteracompute.com/v1","name":"IteraCompute","doc":"https://iteracompute.com/docs.html","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.17,"output":0.7,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.79,"output":2.38,"cache_read":0.026},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (legacy alias)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.17,"output":0.7,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":524288},"cost":{"input":0.3,"output":1.19,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":999999},"cost":{"input":2.35,"output":11.76,"cache_read":0.23},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.69,"output":3.29,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.95,"output":5.7,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.82,"output":2.88,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.08,"output":0.28,"cache_read":0.023},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"opencode":{"id":"opencode","env":["OPENCODE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://opencode.ai/inference/openai/v1","name":"OpenCode","doc":"https://opencode.ai/v2/docs/console/models/","models":{"grok-code":{"id":"grok-code","name":"Grok Code Fast 1","description":"Legacy model retained for compatibility with older integrations","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"xai/grok-code-fast-1"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"kimi-k2.5-free":{"id":"kimi-k2.5-free","name":"Kimi K2.5 Free","description":"Legacy model retained for compatibility with older integrations","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"deepseek-v4.1-flash-free":{"id":"deepseek-v4.1-flash-free","name":"DeepSeek V4.1 Flash Free","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"fledge-alpha-free":{"id":"fledge-alpha-free","name":"Fledge Alpha Free","description":"Free-tier preview reasoning model with text and image input and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"mimo-v2-flash-free":{"id":"mimo-v2-flash-free","name":"MiMo V2 Flash Free","description":"Legacy model retained for compatibility with older integrations","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"minimax-m2.5-free":{"id":"minimax-m2.5-free","name":"MiniMax-M2.5 Free","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.5"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.5"},"ling-3.0-flash-free":{"id":"ling-3.0-flash-free","name":"Ling-3.0-flash Free","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.0-flash"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"flint-40b-free":{"id":"flint-40b-free","name":"Flint 40B Free","description":"Open-weight mixture-of-experts model with about 2.5B active parameters for local coding agents","family":"flint","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06","release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"recursive/flint"},"glm-5.3-flash-free":{"id":"glm-5.3-flash-free","name":"GLM-5.3-Flash Free","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"kimi-k2":{"id":"kimi-k2","name":"Kimi K2","description":"Legacy model retained for compatibility with older integrations","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.4,"output":2.5,"cache_read":0.4},"canonical_model_id":"moonshotai/kimi-k2-0905"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"nemotron-3.5-lightning-free":{"id":"nemotron-3.5-lightning-free","name":"Nemotron 3.5 Lightning Free","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"gemini-3-pro":{"id":"gemini-3-pro","name":"Gemini 3 Pro","description":"Legacy model retained for compatibility with older integrations","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3-pro-preview"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"qwen3.6-plus-free":{"id":"qwen3.6-plus-free","name":"Qwen3.6 Plus Free","description":"Legacy model retained for compatibility with older integrations","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.6-plus"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"cat-jam-free":{"id":"cat-jam-free","name":"Cat Jam Free","description":"A text-only reasoning model for coding and agentic workflows, with tool calling and structured JSON outputs.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"glm-4.7-free":{"id":"glm-4.7-free","name":"GLM-4.7 Free","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-4.7"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"hy3-preview-free":{"id":"hy3-preview-free","name":"Hy3 preview Free","description":"Legacy model retained for compatibility with older integrations","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"tencent/hy3-preview"},"minimax-m2.1-free":{"id":"minimax-m2.1-free","name":"MiniMax-M2.1 Free","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.1"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.6-flash"},"gemini-3-flash":{"id":"gemini-3-flash","name":"Gemini 3 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"mimo-v2-omni-free":{"id":"mimo-v2-omni-free","name":"MiMo V2 Omni Free","description":"Legacy model retained for compatibility with older integrations","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2-omni"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"x-preview-f-free":{"id":"x-preview-f-free","name":"Ox Alpha Free (Unlimited)","description":"Stealth reasoning model for coding, agentic tasks, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-08-21","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.6,"output":2.2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.7"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"exo-free":{"id":"exo-free","name":"Exo Free","description":"Free-tier preview reasoning model with text and image input and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"minimax-m2.1":{"id":"minimax-m2.1","name":"MiniMax-M2.1","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.1},"canonical_model_id":"minimax/MiniMax-M2.1"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"trinity-large-preview-free":{"id":"trinity-large-preview-free","name":"Trinity Large Preview","description":"Legacy model retained for compatibility with older integrations","family":"trinity","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-01-27","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0},"canonical_model_id":"arcee-ai/trinity-large-preview"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"ling-3.1-flash-free":{"id":"ling-3.1-flash-free","name":"Ling 3.1 Flash Free","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.1-flash"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":30,"output":180,"cache_read":30},"canonical_model_id":"openai/gpt-5.4-pro"},"gpt-5.3-codex-spark":{"id":"gpt-5.3-codex-spark","name":"GPT-5.3 Codex Spark","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-spark","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex-spark"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":30,"output":180,"cache_read":30},"canonical_model_id":"openai/gpt-5.5-pro"},"mimo-v2-pro-free":{"id":"mimo-v2-pro-free","name":"MiMo V2 Pro Free","description":"Legacy model retained for compatibility with older integrations","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2-pro"},"muse-spark-1.3-contributor-free":{"id":"muse-spark-1.3-contributor-free","name":"Muse Spark 1.3 Free","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for coding and agentic workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meta/muse-spark-1.3"},"longcat-2.5-preview-free":{"id":"longcat-2.5-preview-free","name":"LongCat 2.5 Preview Free","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meituan/longcat-2.5-preview"},"glm-5-free":{"id":"glm-5-free","name":"GLM-5 Free","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5"},"ling-3.0-flash-fin-free":{"id":"ling-3.0-flash-fin-free","name":"Ling 3.0 Flash Fin Free","description":"Finance-enhanced model for financial research, multi-step investment workflows, and long-horizon planning and execution","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-09-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.0-flash-fin"},"nemotron-3-super-free":{"id":"nemotron-3-super-free","name":"Nemotron 3 Super Free","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"provider":{"npm":"@ai-sdk/mistral","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"claude-sonnet-4":{"id":"claude-sonnet-4","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":128000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5.1"},"hy3-free":{"id":"hy3-free","name":"Hy3 Free","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":190000,"input":192000,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"tencent/hy3"},"nemotron-3.5-super-vl-free":{"id":"nemotron-3.5-super-vl-free","name":"Nemotron 3.5 Super VL Free","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"alibaba/qwen3.5-plus"},"space-bunny-free":{"id":"space-bunny-free","name":"Space Bunny Free","description":"Anonymous preview reasoning model for coding, agentic tasks, tool use, and multimodal input","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":524288,"output":524288},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"big-pickle":{"id":"big-pickle","name":"Big Pickle","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"big-pickle","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-10-17","last_updated":"2025-10-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":160000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"minimax-m3-free":{"id":"minimax-m3-free","name":"MiniMax-M3 Free","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"step-5-preview-free":{"id":"step-5-preview-free","name":"Step 5 Preview Free","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"stepfun/step-5-preview"},"ll-alpha-free":{"id":"ll-alpha-free","name":"LL Alpha Free","description":"Best in class local model","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"mimo-v2.5-free":{"id":"mimo-v2.5-free","name":"MiMo V2.5 Free","description":"MiMo omni model for text, image, video, audio, and agents","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"minimax-m3.1-flash-free":{"id":"minimax-m3.1-flash-free","name":"MiniMax M3.1 Flash Free","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"nemotron-3-ultra-free":{"id":"nemotron-3-ultra-free","name":"Nemotron 3 Ultra Free","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"muse-spark-1.2-contributor-free":{"id":"muse-spark-1.2-contributor-free","name":"Muse Spark 1.2 Free","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meta/muse-spark-1.2"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.6,"output":2.2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"deepseek-v4-flash-free":{"id":"deepseek-v4-flash-free","name":"DeepSeek V4 Flash Free","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"laguna-s-2.1-free":{"id":"laguna-s-2.1-free","name":"Laguna S 2.1 Free","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"poolside/laguna-s-2.1"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Legacy model retained for compatibility with older integrations","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.4,"output":2.5,"cache_read":0.4},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5.1-codex"},"beam-free":{"id":"beam-free","name":"Beam Free","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"ling-2.6-flash-free":{"id":"ling-2.6-flash-free","name":"Ling 2.6 Flash Free","description":"Legacy model retained for compatibility with older integrations","family":"ling-flash-free","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":32800},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"mimo-v2.6-flash-free":{"id":"mimo-v2.6-flash-free","name":"MiMo-V2.6-Flash Free","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"north-mini-code-free":{"id":"north-mini-code-free","name":"North Mini Code Free","description":"Cohere coding model for practical software engineering and agentic edits","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"ring-2.6-1t-free":{"id":"ring-2.6-1t-free","name":"Ring 2.6 1T Free","description":"Legacy model retained for compatibility with older integrations","family":"ring-1t-free","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-05-08","last_updated":"2026-05-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":66000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-3.1-pro":{"id":"gemini-3.1-pro","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"longcat-2.0-free":{"id":"longcat-2.0-free","name":"LongCat-2.0 Free","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meituan/longcat-2.0"},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5-codex"},"claude-3-5-haiku":{"id":"claude-3-5-haiku","name":"Claude Haiku 3.5","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.8,"output":4,"cache_read":0.08,"cache_write":1},"canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"ling-3.0-tiny-free":{"id":"ling-3.0-tiny-free","name":"Ling-3.0-tiny Free","description":"Compact MoE model for responsive agents, instruction following, and multi-turn conversations","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"qwen3-coder":{"id":"qwen3-coder","name":"Qwen3 Coder","description":"Legacy model retained for compatibility with older integrations","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.45,"output":1.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"cohere":{"id":"cohere","env":["COHERE_API_KEY"],"npm":"@ai-sdk/cohere","name":"Cohere","doc":"https://docs.cohere.com/docs/models","models":{"north-small-translate-09-2026":{"id":"north-small-translate-09-2026","name":"North Small Translate","description":"Open-weight translation model for more than 50 languages and locale variants","family":"north","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16000},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-small-translate-1-0"},"command-a-translate-08-2025":{"id":"command-a-translate-08-2025","name":"Command A Translate","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-translate-08-2025"},"tiny-aya-water":{"id":"tiny-aya-water","name":"Tiny Aya Water","description":"Compact multilingual model specialized for European and Asia-Pacific languages","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-water"},"command-r-plus-08-2024":{"id":"command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"command-a-plus-05-2026":{"id":"command-a-plus-05-2026","name":"Command A Plus","description":"Cohere's stronger command model for multilingual agents and enterprise workflows","family":"command-a","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04-01","release_date":"2026-05-20","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-plus-05-2026"},"c4ai-aya-vision-32b":{"id":"c4ai-aya-vision-32b","name":"Aya Vision 32B","description":"Open multilingual vision model for OCR, visual reasoning, and image question answering","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-04","last_updated":"2025-05-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4000},"canonical_model_id":"cohere/c4ai-aya-vision-32b"},"tiny-aya-earth":{"id":"tiny-aya-earth","name":"Tiny Aya Earth","description":"Compact multilingual model specialized for West Asian and African languages","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-earth"},"c4ai-aya-expanse-32b":{"id":"c4ai-aya-expanse-32b","name":"Aya Expanse 32B","description":"Open multilingual model optimized for generation across 23 languages","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-10-24","last_updated":"2024-10-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"cohere/c4ai-aya-expanse-32b"},"command-a-reasoning-08-2025":{"id":"command-a-reasoning-08-2025","name":"Command A Reasoning","description":"Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-reasoning-08-2025"},"command-r7b-12-2024":{"id":"command-r7b-12-2024","name":"Command R7B","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"command-a-vision-07-2025":{"id":"command-a-vision-07-2025","name":"Command A Vision","description":"Cohere vision model for multilingual document analysis, OCR, and image understanding","family":"command-a","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-vision-07-2025"},"tiny-aya-global":{"id":"tiny-aya-global","name":"Tiny Aya Global","description":"Compact multilingual model balanced across 70 languages and regions","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-global"},"tiny-aya-fire":{"id":"tiny-aya-fire","name":"Tiny Aya Fire","description":"Compact multilingual model specialized for South Asian languages","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-fire"},"command-r-08-2024":{"id":"command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"command-a-03-2025":{"id":"command-a-03-2025","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-03-2025"},"north-mini-code-1-0":{"id":"north-mini-code-1-0","name":"North Mini Code","description":"Cohere coding model for practical software engineering and agentic edits","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.cohere.ai/compatibility/v1"},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"command-r7b-arabic-02-2025":{"id":"command-r7b-arabic-02-2025","name":"Command R7B Arabic","description":"Open Command R model optimized for Arabic enterprise chat, RAG, and cultural knowledge","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-02-27","last_updated":"2025-02-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-arabic-02-2025"}}},"snowflake-cortex":{"id":"snowflake-cortex","env":["SNOWFLAKE_ACCOUNT","SNOWFLAKE_CORTEX_PAT"],"npm":"@ai-sdk/openai-compatible","api":"https://${SNOWFLAKE_ACCOUNT}.snowflakecomputing.com/api/v2/cortex/v1","name":"Snowflake Cortex","doc":"https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-rest-api","models":{"openai-gpt-4.1":{"id":"openai-gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai-gpt-5.4":{"id":"openai-gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"openai-gpt-5":{"id":"openai-gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5"},"mistral-large2":{"id":"mistral-large2","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/mistral-large-2407"},"openai-gpt-5.1":{"id":"openai-gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1"},"openai-gpt-5-nano":{"id":"openai-gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":0.06,"output":0.44,"cache_read":0.01},"canonical_model_id":"openai/gpt-5-nano"},"openai-gpt-5-mini":{"id":"openai-gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5-mini"},"openai-gpt-5.5":{"id":"openai-gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"beta","cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai-gpt-5.6-terra":{"id":"openai-gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"openai-gpt-5.2":{"id":"openai-gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.2"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"snowflake-llama3.3-70b":{"id":"snowflake-llama3.3-70b","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"openai-gpt-5.6-luna":{"id":"openai-gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-3.1-pro":{"id":"gemini-3.1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"beta","canonical_model_id":"google/gemini-3.1-pro-preview"},"deepseek-r1":{"id":"deepseek-r1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"canonical_model_id":"deepseek/deepseek-r1"},"openai-gpt-5.6-sol":{"id":"openai-gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"}}},"cline-pass":{"id":"cline-pass","env":["CLINE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.cline.bot/api/v1","name":"ClinePass","doc":"https://docs.cline.bot/getting-started/clinepass","models":{"cline-pass/deepseek-v4.1-flash":{"id":"cline-pass/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"cline-pass/glm-5.3":{"id":"cline-pass/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"cline-pass/kimi-k3":{"id":"cline-pass/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"cline-pass/muse-spark-1.3-contributor":{"id":"cline-pass/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"cline-pass/deepseek-v4-pro":{"id":"cline-pass/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"cline-pass/qwen3.7-max":{"id":"cline-pass/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"cline-pass/mimo-v2.6-flash":{"id":"cline-pass/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"cline-pass/kimi-k2.7-code":{"id":"cline-pass/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"cline-pass/mimo-v2.6-pro":{"id":"cline-pass/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"cline-pass/deepseek-v4-flash":{"id":"cline-pass/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"cline-pass/mimo-v2.5":{"id":"cline-pass/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"cline-pass/qwen3.8-max":{"id":"cline-pass/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"cline-pass/kimi-k2.6":{"id":"cline-pass/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"cline-pass/minimax-m3":{"id":"cline-pass/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"cline-pass/mimo-v2.5-pro":{"id":"cline-pass/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.74,"output":3.48,"cache_read":0.0145},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"cline-pass/glm-5.2":{"id":"cline-pass/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"cline-pass/glm-5.3-flash":{"id":"cline-pass/glm-5.3-flash","name":"cline-pass/glm-5.3-flash","description":"Latest natively multimodal model in the GLM-5 series","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"cline-pass/qwen3.7-plus":{"id":"cline-pass/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"}}},"fastrouter":{"id":"fastrouter","env":["FASTROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://go.fastrouter.ai/api/v1","name":"FastRouter","doc":"https://fastrouter.ai/models","models":{"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"sarvam/sarvam-105b":{"id":"sarvam/sarvam-105b","name":"Sarvam 105B","description":"Flagship Indian-language reasoning model for enterprise multilingual applications","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.04,"output":0.16,"cache_read":0.025},"canonical_model_id":"sarvam/sarvam-105b"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4":{"id":"x-ai/grok-4","name":"Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.75,"cache_write":15}},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"google/veo3.1-lite":{"id":"google/veo3.1-lite","type":"video","name":"Veo 3.1 Lite","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"google/veo-3.1-lite-generate-preview"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/veo3.1":{"id":"google/veo3.1","type":"video","name":"Veo 3.1","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"google/veo-3.1-generate-preview"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/veo3.1-fast":{"id":"google/veo3.1-fast","type":"video","name":"Veo 3.1 Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-01-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"google/veo-3.1-fast-generate-preview"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.25,"cache_read":0.007},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-image-2":{"id":"openai/gpt-image-2","type":"image","name":"GPT Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"openai/gpt-image-2"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-10-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.09,"output":0.36},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-10-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-10-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"bytedance/seedance-2":{"id":"bytedance/seedance-2","type":"video","name":"Seedance 2","description":"ByteDance unified multimodal audio-video model supporting text, image, video and audio references, editing, extension and synchronized stereo sound.","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-12","last_updated":"2026-10-04","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":4096,"output":0},"canonical_model_id":"bytedance-seed/seedance-2.0"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.75,"output":3.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"wanx/wan-v2-6":{"id":"wanx/wan-v2-6","type":"video","name":"Wan 2.6","description":"Video model for text- and image-to-video with audio input and synchronized sound","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-12-03","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"alibaba/wan2.6"},"leonardo-ai/lucid-origin":{"id":"leonardo-ai/lucid-origin","name":"Lucid Origin","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"lucid","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-01","last_updated":"2025-06-01","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":4096,"output":0}},"leonardo-ai/lucid-realism":{"id":"leonardo-ai/lucid-realism","name":"Lucid Realism","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"lucid","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-01","last_updated":"2025-06-01","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":4096,"output":0}},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.08,"cache_read":0.12},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.05,"output":3.5,"cache_read":0.205},"canonical_model_id":"zhipuai/glm-5.1"}}},"vercel":{"id":"vercel","env":["AI_GATEWAY_API_KEY"],"npm":"@ai-sdk/gateway","name":"Vercel AI Gateway","doc":"https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway","models":{"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":80000},"cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.075,"output":0.3,"cache_read":0.0015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v3.2-thinking":{"id":"deepseek/deepseek-v3.2-thinking","name":"DeepSeek V3.2 Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":0.62,"output":1.85},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4.1-flash-fast":{"id":"deepseek/deepseek-v4.1-flash-fast","name":"DeepSeek V4.1 Flash Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.075,"output":0.3,"cache_read":0.0015},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.066},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.0068},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.076,"output":0.153,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek-V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":128000},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek/deepseek-v4.1-flash-1017-fast":{"id":"deepseek/deepseek-v4.1-flash-1017-fast","name":"DeepSeek V4.1 Flash 1017 Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":0.62,"output":1.85},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-r1":{"id":"deepseek/deepseek-r1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/namazu":{"id":"sakana/namazu","name":"Sakana Namazu","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"voyage/voyage-3.5-lite":{"id":"voyage/voyage-3.5-lite","type":"embedding","name":"voyage-3.5-lite","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-finance-2":{"id":"voyage/voyage-finance-2","type":"embedding","name":"voyage-finance-2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-06-03","last_updated":"2024-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/rerank-3":{"id":"voyage/rerank-3","type":"reranking","name":"Voyage Rerank 3","description":"Reranker optimized for retrieval quality, successor to rerank-2.5","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-3"},"voyage/voyage-law-2":{"id":"voyage/voyage-law-2","type":"embedding","name":"voyage-law-2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-04-15","last_updated":"2024-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/rerank-3-lite":{"id":"voyage/rerank-3-lite","type":"reranking","name":"Voyage Rerank 3 Lite","description":"Reranker optimized for latency and cost, successor to rerank-2.5-lite","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-3-lite"},"voyage/voyage-3.5":{"id":"voyage/voyage-3.5","type":"embedding","name":"voyage-3.5","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-code-3":{"id":"voyage/voyage-code-3","type":"embedding","name":"voyage-code-3","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-12-04","last_updated":"2024-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/rerank-2.5":{"id":"voyage/rerank-2.5","type":"reranking","name":"Voyage Rerank 2.5","description":"Instruction-following reranker optimized for retrieval quality","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-2.5"},"voyage/voyage-code-2":{"id":"voyage/voyage-code-2","type":"embedding","name":"voyage-code-2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-4-lite":{"id":"voyage/voyage-4-lite","type":"embedding","name":"voyage-4-lite","description":"Lightweight general-purpose embedding model optimized for latency and cost","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/voyage-4-lite"},"voyage/rerank-2.5-lite":{"id":"voyage/rerank-2.5-lite","type":"reranking","name":"Voyage Rerank 2.5 Lite","description":"Instruction-following reranker optimized for latency and quality","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-2.5-lite"},"voyage/voyage-4":{"id":"voyage/voyage-4","type":"embedding","name":"voyage-4","description":"General-purpose and multilingual embedding model balancing retrieval quality and cost","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/voyage-4"},"voyage/voyage-3-large":{"id":"voyage/voyage-3-large","type":"embedding","name":"voyage-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-01-07","last_updated":"2024-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-4-large":{"id":"voyage/voyage-4-large","type":"embedding","name":"voyage-4-large","description":"General-purpose and multilingual embedding model optimized for retrieval quality","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/voyage-4-large"},"spacexai/grok-4.20-multi-agent-beta":{"id":"spacexai/grok-4.20-multi-agent-beta","name":"Grok 4.20 Multi Agent Beta","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-tts":{"id":"spacexai/grok-tts","type":"speech","name":"Grok TTS","description":"Convert text into spoken audio with a single API call. The API supports a rich set of expressive voices, inline speech tags for fine-grained delivery control, and output formats from high-fidelity MP3 to telephony-optimized μ-law.","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":15000,"output":15000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-tts"},"spacexai/grok-4.3":{"id":"spacexai/grok-4.3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"spacexai/grok-voice-think-fast-2.0":{"id":"spacexai/grok-voice-think-fast-2.0","type":"realtime","name":"Grok Voice Think Fast 2.0","description":"Realtime voice model for low-latency spoken conversations","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"spacexai/grok-imagine-video-1.5":{"id":"spacexai/grok-imagine-video-1.5","type":"video","name":"Grok Imagine Video 1.5","description":"Video model for text-to-video, image-to-video, and reference-to-video generation with optional voice references","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-30","last_updated":"2026-05-30","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-video-1.5"},"spacexai/grok-4.20-non-reasoning-beta":{"id":"spacexai/grok-4.20-non-reasoning-beta","name":"Grok 4.20 Beta Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.4,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-stt":{"id":"spacexai/grok-stt","type":"transcription","name":"Grok STT","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"spacexai/grok-4.6":{"id":"spacexai/grok-4.6","name":"Grok 4.6","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"spacexai/grok-4.20-multi-agent":{"id":"spacexai/grok-4.20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"release_date":"2026-03-10","last_updated":"2026-03-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-build-0.1":{"id":"spacexai/grok-build-0.1","name":"Grok Build 0.1","description":"Grok coding model for agentic engineering, edits, and codebase workflows","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"spacexai/grok-4.20-reasoning-beta":{"id":"spacexai/grok-4.20-reasoning-beta","name":"Grok 4.20 Beta Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-imagine-image":{"id":"spacexai/grok-imagine-image","type":"image","name":"Grok Imagine Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-image"},"spacexai/grok-4.20-reasoning":{"id":"spacexai/grok-4.20-reasoning","name":"Grok 4.20 Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"spacexai/grok-imagine-image-2.0":{"id":"spacexai/grok-imagine-image-2.0","type":"image","name":"Grok Imagine Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":64000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-image-2.0"},"spacexai/grok-4.7":{"id":"spacexai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"spacexai/grok-imagine-video-1.5-lite":{"id":"spacexai/grok-imagine-video-1.5-lite","type":"video","name":"Grok Imagine Video 1.5 Lite","description":"Cost-efficient video model for text-to-video and image-to-video generation","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-video-1.5-lite"},"spacexai/grok-imagine-video":{"id":"spacexai/grok-imagine-video","type":"video","name":"Grok Imagine","description":"Video model for text-to-video, image-to-video, video editing and extension with native synchronized audio","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-video"},"spacexai/grok-4.20-non-reasoning":{"id":"spacexai/grok-4.20-non-reasoning","name":"Grok 4.20 Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"spacexai/grok-4.5":{"id":"spacexai/grok-4.5","name":"Grok 4.5","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"inception/mercury-3":{"id":"inception/mercury-3","name":"Mercury 3","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01}},"inception/mercury-coder-small":{"id":"inception/mercury-coder-small","name":"Mercury Coder Small Beta","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"mercury","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":16384},"cost":{"input":0.25,"output":1}},"inception/mercury-2":{"id":"inception/mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.04,"output":0.15,"cache_read":0.004},"canonical_model_id":"inception/mercury-2.5"},"scaledown/scaledown-extract-1":{"id":"scaledown/scaledown-extract-1","name":"Extract-1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":0.05,"output":0}},"scaledown/scaledown-summarize-1":{"id":"scaledown/scaledown-summarize-1","name":"Summarize-1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000},"cost":{"input":0.05,"output":0}},"topaz/proteus":{"id":"topaz/proteus","type":"video","name":"Proteus","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"topaz/starlight-precise-2.6":{"id":"topaz/starlight-precise-2.6","type":"video","name":"Starlight Precise 2.6","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"topaz/wonder-3.5":{"id":"topaz/wonder-3.5","type":"image","name":"Wonder 3.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"Morph v3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.9,"output":1.9}},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"Morph v3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":0.8,"output":1.2}},"recraft/recraft-v4.1":{"id":"recraft/recraft-v4.1","type":"image","name":"Recraft V4.1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4":{"id":"recraft/recraft-v4","type":"image","name":"Recraft V4","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1-utility":{"id":"recraft/recraft-v4.1-utility","type":"image","name":"Recraft V4.1 Utility","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1-pro":{"id":"recraft/recraft-v4.1-pro","type":"image","name":"Recraft V4.1 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1-utility-pro":{"id":"recraft/recraft-v4.1-utility-pro","type":"image","name":"Recraft V4.1 Utility Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4-pro":{"id":"recraft/recraft-v4-pro","type":"image","name":"Recraft V4 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v2":{"id":"recraft/recraft-v2","type":"image","name":"Recraft V2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-13","last_updated":"2024-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1-flash":{"id":"recraft/recraft-v4.1-flash","type":"image","name":"Recraft V4.1 Flash","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v3":{"id":"recraft/recraft-v3","type":"image","name":"Recraft V3","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-30","last_updated":"2024-10","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"reflection/beam":{"id":"reflection/beam","name":"Beam 1.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"reflection/beam-free":{"id":"reflection/beam-free","name":"Beam 1.0 (Free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131100},"cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.2,"output":4,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5.3-flashx":{"id":"zai/glm-5.3-flashx","name":"GLM 5.3 FlashX","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":64000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":66000,"output":16000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.7-flash":{"id":"zai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131000},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai/glm-5.2-fast":{"id":"zai/glm-5.2-fast","name":"GLM 5.2 Fast","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":2.8,"output":8.8,"cache_read":0.56},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM 4.7 FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":120000},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-5.3-fast":{"id":"zai/glm-5.3-fast","name":"GLM 5.3 Fast","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.8,"output":2.55,"cache_read":0.16},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-5v-turbo":{"id":"zai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":96000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.2,"output":1.1,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-4.5-air"},"alibaba/qwen3.8-27b":{"id":"alibaba/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.8-27b"},"alibaba/qwen-3-32b":{"id":"alibaba/qwen-3-32b","name":"Qwen 3.32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-32b"},"alibaba/qwen3.8-max-prime":{"id":"alibaba/qwen3.8-max-prime","name":"Qwen 3.8 Max Prime","description":"High-throughput edition of Qwen3.8 Max for coding, professional work, multimodal understanding, and long-running agent workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5,"cache_write":5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"alibaba/wan-v2.6-r2v-flash":{"id":"alibaba/wan-v2.6-r2v-flash","type":"video","name":"Wan v2.6 Reference-to-Video Flash","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3-vl-instruct":{"id":"alibaba/qwen3-vl-instruct","name":"Qwen3 VL Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.88,"cache_read":0.11}},"alibaba/qwen3.6-plus":{"id":"alibaba/qwen3.6-plus","name":"Qwen 3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":131072}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"alibaba/qwen-3-30b":{"id":"alibaba/qwen-3-30b","name":"Qwen3-30B-A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"alibaba/qwen3.7-flash":{"id":"alibaba/qwen3.7-flash","name":"Qwen 3.7 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":64000},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"alibaba/qwen3.7-max":{"id":"alibaba/qwen3.7-max","name":"Qwen 3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":64000},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"alibaba/qwen3-embedding-0.6b":{"id":"alibaba/qwen3-embedding-0.6b","type":"embedding","name":"Qwen3 Embedding 0.6B","description":"Compact multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"alibaba/qwen3-embedding-0.6b"},"alibaba/qwen3-next-80b-a3b-thinking":{"id":"alibaba/qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B Thinking","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"alibaba/qwen3-embedding-8b":{"id":"alibaba/qwen3-embedding-8b","type":"embedding","name":"Qwen3 Embedding 8B","description":"Multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"alibaba/wan-v2.5-t2v-preview":{"id":"alibaba/wan-v2.5-t2v-preview","type":"video","name":"Wan v2.5 Text-to-Video Preview","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-24","last_updated":"2025-09-24","modalities":{"input":["text","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3-embedding-4b":{"id":"alibaba/qwen3-embedding-4b","type":"embedding","name":"Qwen3 Embedding 4B","description":"Multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"alibaba/qwen3-embedding-4b"},"alibaba/wan-v3.0-video-prime":{"id":"alibaba/wan-v3.0-video-prime","type":"video","name":"Wan v3.0 Video Prime","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.8-omni-flash":{"id":"alibaba/qwen3.8-omni-flash","name":"Qwen 3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"alibaba/qwen-3-14b":{"id":"alibaba/qwen-3-14b","name":"Qwen3-14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.24}},"alibaba/wan-v3.0-video":{"id":"alibaba/wan-v3.0-video","type":"video","name":"Wan v3.0 Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-23","last_updated":"2026-08-23","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen-3-235b":{"id":"alibaba/qwen-3-235b","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.22,"output":0.88},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"alibaba/wan-v2.6-i2v-flash":{"id":"alibaba/wan-v2.6-i2v-flash","type":"video","name":"Wan v2.6 Image-to-Video Flash","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen 3.8 Max","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"alibaba/qwen3-max-thinking":{"id":"alibaba/qwen3-max-thinking","name":"Qwen 3 Max Thinking","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-23","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":1.2,"output":6,"cache_read":0.24,"tiers":[{"input":2.4,"output":12,"cache_read":0.48,"tier":{"type":"context","size":32001}},{"input":3,"output":15,"cache_read":0.6,"tier":{"type":"context","size":128001}}]}},"alibaba/qwen3.8-max-0902":{"id":"alibaba/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"alibaba/qwen3.8-2.4t-a95b":{"id":"alibaba/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"alibaba/qwen3.5-plus":{"id":"alibaba/qwen3.5-plus","name":"Qwen 3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":256001}}],"context_over_200k":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625}},"canonical_model_id":"alibaba/qwen3.5-plus"},"alibaba/wan-v2.7-r2v":{"id":"alibaba/wan-v2.7-r2v","type":"video","name":"Wan v2.7 Reference-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.6-27b":{"id":"alibaba/qwen3.6-27b","name":"Qwen 3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.6-27b"},"alibaba/qwen3.8-flash":{"id":"alibaba/qwen3.8-flash","name":"Qwen 3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":128000},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"alibaba/qwen3.7-plus":{"id":"alibaba/qwen3.7-plus","name":"Qwen 3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"},"alibaba/wan-v2.6-i2v":{"id":"alibaba/wan-v2.6-i2v","type":"video","name":"Wan v2.6 Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/wan-v2.6-t2v":{"id":"alibaba/wan-v2.6-t2v","type":"video","name":"Wan v2.6 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/wan-v2.7-t2v":{"id":"alibaba/wan-v2.7-t2v","type":"video","name":"Wan v2.7 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.5-flash":{"id":"alibaba/qwen3.5-flash","name":"Qwen 3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"alibaba/qwen3-coder-30b-a3b":{"id":"alibaba/qwen3-coder-30b-a3b","name":"Qwen 3 Coder 30B A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"alibaba/qwen3-coder-next":{"id":"alibaba/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"alibaba/qwen3-next-80b-a3b-instruct":{"id":"alibaba/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262114,"output":262114},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"alibaba/wan-v2.6-r2v":{"id":"alibaba/wan-v2.6-r2v","type":"video","name":"Wan v2.6 Reference-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3-coder":{"id":"alibaba/qwen3-coder","name":"Qwen3 Coder 480B A35B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.22,"output":1.8,"cache_read":0.022,"tiers":[{"input":2.7,"output":13.5,"cache_read":0.54,"tier":{"type":"context","size":32001}},{"input":4.5,"output":22.5,"cache_read":0.9,"tier":{"type":"context","size":128001}}]},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"heygen/heygen-video-1":{"id":"heygen/heygen-video-1","type":"video","name":"HeyGen Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-09","last_updated":"2026-10-09","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"inclusionai/ling-3.1-flash"},"inclusionai/ling-3.0-flash-sante":{"id":"inclusionai/ling-3.0-flash-sante","name":"Ling 3.0 Flash Sante","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-sante"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.021,"output":0.063,"cache_read":0.0042},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.0-flash-fin":{"id":"inclusionai/ling-3.0-flash-fin","name":"Ling 3.0 Flash Fin","description":"Finance-enhanced model for financial research, multi-step investment workflows, and long-horizon planning and execution","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-09-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-fin"},"quiverai/arrow-2":{"id":"quiverai/arrow-2","name":"Arrow 2","description":"Fast SVG generation model for creation, vectorization, editing, and animation","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"release_date":"2026-09-16","last_updated":"2026-09-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"quiverai/arrow-2"},"quiverai/arrow-2-telos":{"id":"quiverai/arrow-2-telos","name":"Arrow 2 Telos","description":"High-fidelity SVG generation model for complex vector work and long-context refinement","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"release_date":"2026-09-16","last_updated":"2026-09-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":6,"output":30,"cache_read":0.6,"cache_write":7.5},"canonical_model_id":"quiverai/arrow-2-telos"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131000},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"Minimax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax M2.5 High Speed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131000},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.1-lightning":{"id":"minimax/minimax-m2.1-lightning","name":"MiniMax M2.1 Lightning","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-12-23","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":2.4,"cache_read":0.03,"cache_write":0.375}},"minimax/minimax-h3":{"id":"minimax/minimax-h3","type":"video","name":"MiniMax H3","description":"MiniMax general-purpose video model with multimodal references and synchronized stereo audio. Open H3-Base checkpoints; Context-IR and 2K regeneration are not included in the initial weights release.","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-31","last_updated":"2026-10-04","modalities":{"input":["text","image"],"output":["video"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"minimax/MiniMax-H3"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":196608},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 High Speed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-h3-max":{"id":"minimax/minimax-h3-max","type":"video","name":"MiniMax H3 Max","description":"Fast video model post-trained by fal.ai on MiniMax H3, with synchronized audio at 480p and 768p","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"minimax/MiniMax-H3-Max"},"cohere/embed-v5.0-pro":{"id":"cohere/embed-v5.0-pro","type":"embedding","name":"Embed 5 Pro","description":"Cohere's most capable multimodal embedding model for text, images, and mixed documents","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":2048},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/embed-v5.0-pro"},"cohere/rerank-v4-pro":{"id":"cohere/rerank-v4-pro","type":"reranking","name":"Cohere Rerank 4 Pro","description":"Multilingual reranker for maximum relevance on documents and semi-structured data","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/rerank-v4.0-pro"},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a"},"cohere/rerank-v3.5":{"id":"cohere/rerank-v3.5","type":"reranking","name":"Cohere Rerank 3.5","description":"Multilingual reranker for documents and semi-structured data","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-12-02","last_updated":"2024-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/rerank-v3.5"},"cohere/embed-v4.0":{"id":"cohere/embed-v4.0","type":"embedding","name":"Embed v4.0","description":"Multimodal embedding model for text, images, and mixed text-and-image documents","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":1536},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/embed-v4.0"},"cohere/rerank-v5-pro":{"id":"cohere/rerank-v5-pro","type":"reranking","name":"Cohere Rerank 5 Pro","description":"Reranking model for improving retrieval quality in search and recommendation systems","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"cohere/rerank-v4-fast":{"id":"cohere/rerank-v4-fast","type":"reranking","name":"Cohere Rerank 4 Fast","description":"Low-latency multilingual reranker for high-throughput retrieval","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/rerank-v4.0-fast"},"cohere/rerank-v5-fast":{"id":"cohere/rerank-v5-fast","type":"reranking","name":"Cohere Rerank 5 Fast","description":"Reranking model for improving retrieval quality in search and recommendation systems","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"cohere/embed-v5.0-fast":{"id":"cohere/embed-v5.0-fast","type":"embedding","name":"Embed 5 Fast","description":"Faster, lighter Embed 5 model sharing Embed 5 Pro's embedding space","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":2048},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/embed-v5.0-fast"},"microsoft/mai-transcribe-2":{"id":"microsoft/mai-transcribe-2","type":"transcription","name":"MAI-Transcribe 2","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mai","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-03","last_updated":"2026-09-03","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-voice-2.1-flash":{"id":"microsoft/mai-voice-2.1-flash","type":"speech","name":"MAI-Voice-2.1-Flash","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-transcribe-2-streaming":{"id":"microsoft/mai-transcribe-2-streaming","type":"transcription","name":"MAI-Transcribe-2-Streaming","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mai","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-voice-2":{"id":"microsoft/mai-voice-2","type":"speech","name":"MAI-Voice-2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-image-2.6-flash":{"id":"microsoft/mai-image-2.6-flash","type":"image","name":"MAI-Image-2.6-Flash","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":1.75,"output":19}},"microsoft/mai-voice-2.1":{"id":"microsoft/mai-voice-2.1","type":"speech","name":"MAI-Voice-2.1","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-voice-2-flash":{"id":"microsoft/mai-voice-2-flash","type":"speech","name":"MAI-Voice-2-Flash","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-image-2.6":{"id":"microsoft/mai-image-2.6","type":"image","name":"MAI-Image-2.6","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":38}},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.09,"output":0.18,"cache_read":0.009},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-s-2.1-free":{"id":"poolside/laguna-s-2.1-free","name":"Laguna S 2.1 Free","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-s-2.1"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo M2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131100},"cost":{"input":0.119,"output":0.238,"cache_read":0.0025},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131000},"cost":{"input":0.3045,"output":0.609,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"google/gemini-3.8-flash-tts":{"id":"google/gemini-3.8-flash-tts","type":"speech","name":"Gemini 3.8 Flash TTS","description":"Flagship text-to-speech model for expressive, studio-grade voices and long-form multi-speaker audio","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.5,"output":9,"output_audio":9},"canonical_model_id":"google/gemini-3.8-flash-tts"},"google/veo-3.0-fast-generate-001":{"id":"google/veo-3.0-fast-generate-001","type":"video","name":"Veo 3.0 Fast Generate","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-07-31","last_updated":"2026-06-08","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-omni-flash-preview":{"id":"google/gemini-omni-flash-preview","type":"video","name":"Gemini Omni Flash Preview","description":"Omni-modal model for text, vision, audio, and multimodal agent tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":57920},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-omni-flash-preview"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-embedding-001":{"id":"google/gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"google/gemini-embedding-001"},"google/gemini-3.8-live":{"id":"google/gemini-3.8-live","type":"realtime","name":"Gemini 3.8 Live","description":"Low-latency Live API model for real-time voice agents with interleaved reasoning and async tools","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-09-15","last_updated":"2026-09-15","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.75,"output":4.5,"input_audio":3,"output_audio":12},"canonical_model_id":"google/gemini-3.8-live"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-transcribe":{"id":"google/gemini-3.5-transcribe","type":"transcription","name":"Gemini 3.5 Transcribe","description":"Speech-to-text model with language detection, speaker diarization, word timestamps, and custom vocabulary","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":98304,"output":32768},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":2,"output":12,"input_audio":2},"canonical_model_id":"google/gemini-3.5-transcribe"},"google/gemini-3-flash":{"id":"google/gemini-3-flash","name":"Gemini 3 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.5,"output":3,"cache_read":0.05}},"google/veo-3.1-lite-generate-001":{"id":"google/veo-3.1-lite-generate-001","type":"video","name":"Veo 3.1 Lite Generate","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.8-flash-lite-tts":{"id":"google/gemini-3.8-flash-lite-tts","type":"speech","name":"Gemini 3.8 Flash-Lite TTS","description":"Fast, cost-efficient text-to-speech model for high-volume narration and voice agent pipelines","family":"gemini-flash-lite","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.5,"output":6,"output_audio":6},"canonical_model_id":"google/gemini-3.8-flash-lite-tts"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/text-multilingual-embedding-002":{"id":"google/text-multilingual-embedding-002","type":"embedding","name":"Text Multilingual Embedding 002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-01","last_updated":"2024-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Gemini Nano Banana 2.1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/veo-3.1-generate-001":{"id":"google/veo-3.1-generate-001","type":"video","name":"Veo 3.1","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-06-08","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3.5-transcribe-live":{"id":"google/gemini-3.5-transcribe-live","type":"transcription","name":"Gemini 3.5 Transcribe Live","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"google/gemini-3.5-transcribe-live"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/veo-3.1-fast-generate-001":{"id":"google/veo-3.1-fast-generate-001","type":"video","name":"Veo 3.1 Fast Generate","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-06-08","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/veo-3.0-generate-001":{"id":"google/veo-3.0-generate-001","type":"video","name":"Veo 3.0","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-20","last_updated":"2026-06-08","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite)","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Gemini 3.1 Flash Image Preview (Nano Banana 2)","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"google/text-embedding-005":{"id":"google/text-embedding-005","type":"embedding","name":"Text Embedding 005","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-01","last_updated":"2024-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana (Gemini 2.5 Flash Image)","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-3.8-live-extended-thinking":{"id":"google/gemini-3.8-live-extended-thinking","type":"realtime","name":"Gemini 3.8 Live Extended Thinking","description":"Live API voice model with configurable background reasoning for multi-step real-time agents","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-09-15","last_updated":"2026-09-15","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.75,"output":4.5,"input_audio":3,"output_audio":12},"canonical_model_id":"google/gemini-3.8-live-extended-thinking"},"google/gemini-embedding-2":{"id":"google/gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"google/gemini-embedding-2"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite"},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"mistral/mistral-large-3":{"id":"mistral/mistral-large-3","name":"Mistral Large 3","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/voxtral-small-transcribe-3":{"id":"mistral/voxtral-small-transcribe-3","type":"transcription","name":"Voxtral Small Transcribe 3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"mistral/voxtral-mini-tts-3":{"id":"mistral/voxtral-mini-tts-3","type":"speech","name":"Voxtral TTS 3","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"mistral/codestral":{"id":"mistral/codestral","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-latest"},"mistral/codestral-embed":{"id":"mistral/codestral-embed","type":"embedding","name":"Codestral Embed","description":"Embedding model for code search, retrieval, and code-aware RAG pipelines","family":"codestral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"mistral/codestral-embed-2505"},"mistral/mistral-nemo":{"id":"mistral/mistral-nemo","name":"Mistral Nemo","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.029,"output":0.03},"canonical_model_id":"mistral/mistral-nemo"},"mistral/ministral-8b":{"id":"mistral/ministral-8b","name":"Ministral 8B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-16","last_updated":"2024-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4000},"cost":{"input":0.15,"output":0.15,"cache_read":0.015}},"mistral/mistral-embed":{"id":"mistral/mistral-embed","type":"embedding","name":"Mistral Embed","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":3072},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"mistral/mistral-embed"},"mistral/ministral-14b":{"id":"mistral/ministral-14b","name":"Ministral 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b"},"mistral/ministral-3b":{"id":"mistral/ministral-3b","name":"Ministral 3B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-16","last_updated":"2024-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4000},"cost":{"input":0.1,"output":0.1,"cache_read":0.01}},"mistral/mistral-medium-3.5":{"id":"mistral/mistral-medium-3.5","name":"Mistral Medium Latest","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":256000},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistral/mistral-large-4":{"id":"mistral/mistral-large-4","name":"Mistral Large 4","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistral/voxtral-mini-transcribe-3":{"id":"mistral/voxtral-mini-transcribe-3","type":"transcription","name":"Voxtral Mini Transcribe 3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"mistral/mistral-small":{"id":"mistral/mistral-small","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-latest"},"openai/gpt-5.1-thinking":{"id":"openai/gpt-5.1-thinking","name":"GPT 5.1 Thinking","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-11-12","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"openai/tts-1":{"id":"openai/tts-1","type":"speech","name":"TTS-1","description":"Text-to-speech model optimized for low latency","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/tts-1"},"openai/gpt-6-sol-fast":{"id":"openai/gpt-6-sol-fast","name":"GPT-6 Sol (Fast)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.5-fast":{"id":"openai/gpt-5.5-fast","name":"GPT 5.5 (Fast)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":872000,"output":128000},"cost":{"input":12.5,"output":75,"cache_read":1.25},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.3-codex-fast":{"id":"openai/gpt-5.3-codex-fast","name":"GPT 5.3 Codex (Fast)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":3.5,"output":28,"cache_read":0.35},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT 5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/o3-fast":{"id":"openai/o3-fast","name":"o3 (Fast)","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"cost":{"input":3.5,"output":14,"cache_read":0.875},"canonical_model_id":"openai/o3"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT 5.2 ","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-oss-safeguard-120b":{"id":"openai/gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":112000,"output":16000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2-Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-5.6-terra-fast":{"id":"openai/gpt-5.6-terra-fast","name":"GPT 5.6 Terra (Fast)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":36,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":36,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-image-2.5-flare":{"id":"openai/gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"openai/gpt-5.6-luna-fast":{"id":"openai/gpt-5.6-luna-fast","name":"GPT 5.6 Luna (Fast)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":0.8,"output":3.6,"cache_read":0.08,"cache_write":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.8,"output":3.6,"cache_read":0.08,"cache_write":1}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-realtime-mini":{"id":"openai/gpt-realtime-mini","type":"realtime","name":"GPT-Realtime mini","description":"Cost-efficient realtime speech-to-speech model for voice agents","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32000,"output":4096},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"input_audio":10,"output_audio":20},"canonical_model_id":"openai/gpt-realtime-mini"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-realtime-2":{"id":"openai/gpt-realtime-2","type":"realtime","name":"gpt-realtime-2","description":"Realtime speech-to-speech model with configurable reasoning and reliable tool use for voice agents","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":32000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":4,"output":24,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-2"},"openai/gpt-image-2.5-sunburst":{"id":"openai/gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":122880,"output":8192},"cost":{"input":0.03,"output":0.14},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-image-1-mini":{"id":"openai/gpt-image-1-mini","type":"image","name":"GPT Image 1 Mini","description":"Cost-efficient GPT Image 1 variant for image generation and editing","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":2,"output":8,"cache_read":0.2},"canonical_model_id":"openai/gpt-image-1-mini"},"openai/gpt-5.4-mini-fast":{"id":"openai/gpt-5.4-mini-fast","name":"GPT 5.4 Mini (Fast)","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"gpt-oss-safeguard-20b","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":112000,"output":16000},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":0.2,"output":0.8,"cache_read":0.06},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.1,"output":0.4,"cache_read":0.03},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-4o-fast":{"id":"openai/gpt-4o-fast","name":"GPT-4o (Fast)","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"cost":{"input":4.25,"output":17,"cache_read":2.125},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-6.1-sol-fast":{"id":"openai/gpt-6.1-sol-fast","name":"GPT-6.1 Sol (Fast)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.4,"cache_write":10,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.4,"cache_write":10}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT 5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT 5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-image-2":{"id":"openai/gpt-image-2","type":"image","name":"GPT Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT 5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/text-embedding-3-small":{"id":"openai/text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Small, efficient OpenAI text embedding model","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/text-embedding-3-small"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":0.25,"output":1,"cache_read":0.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT 5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-realtime-1.5":{"id":"openai/gpt-realtime-1.5","type":"realtime","name":"GPT-Realtime-1.5","description":"Realtime speech-to-speech model for voice agents and customer support","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32000,"output":4096},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":4,"output":16,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-1.5"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":0.7,"output":2.8,"cache_read":0.175},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.125,"output":1,"cache_read":0.0125},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT 5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT 5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":872000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-4o-mini-transcribe":{"id":"openai/gpt-4o-mini-transcribe","type":"transcription","name":"GPT-4o mini Transcribe","description":"Speech-to-text model powered by GPT-4o Mini for fast, accurate transcription","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-06-01","release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":1.25,"output":5,"input_audio":1.25},"canonical_model_id":"openai/gpt-4o-mini-transcribe"},"openai/gpt-5-mini-fast":{"id":"openai/gpt-5-mini-fast","name":"GPT-5 mini (Fast)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":128000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-realtime-2.1":{"id":"openai/gpt-realtime-2.1","type":"realtime","name":"gpt-realtime-2.1","description":"Realtime speech-to-speech model with configurable reasoning, tool use, and robust voice-agent behavior","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"input":96000,"output":32000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":4,"output":24,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-2.1"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4o-mini-fast":{"id":"openai/gpt-4o-mini-fast","name":"GPT-4o mini (Fast)","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"cost":{"input":0.25,"output":1,"cache_read":0.125},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-6-astra-fast":{"id":"openai/gpt-6-astra-fast","name":"GPT-6 Astra (Fast)","description":"Fast variant of GPT-6 Astra for low-latency assistance and high-volume workloads.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":20,"output":100,"cache_read":2,"cache_write":25,"tiers":[{"input":40,"output":150,"cache_read":4,"cache_write":50,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":40,"output":150,"cache_read":4,"cache_write":50}},"canonical_model_id":"openai/gpt-6-astra-fast"},"openai/gpt-5.4-fast":{"id":"openai/gpt-5.4-fast","name":"GPT 5.4 (Fast)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-4o-transcribe":{"id":"openai/gpt-4o-transcribe","type":"transcription","name":"GPT-4o Transcribe","description":"Speech-to-text model powered by GPT-4o for accurate transcription","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-06-01","release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":2.5,"output":10,"input_audio":2.5},"canonical_model_id":"openai/gpt-4o-transcribe"},"openai/gpt-5.1-thinking-fast":{"id":"openai/gpt-5.1-thinking-fast","name":"GPT 5.1 Thinking (Fast)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"release_date":"2025-11-12","last_updated":"2025-11-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2.5,"output":20,"cache_read":0.25}},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT 5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":4.25,"output":17,"cache_read":2.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-5.2-fast":{"id":"openai/gpt-5.2-fast","name":"GPT 5.2 (Fast)","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":3.5,"output":28,"cache_read":0.35},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-6-luna-fast":{"id":"openai/gpt-6-luna-fast","name":"GPT-6 Luna (Fast)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.5,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":0.4,"output":1.5,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-6-luna"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":2,"output":8,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.55,"output":2.2,"cache_read":0.1375},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.5,"cache_read":0.1},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-4.1-mini-fast":{"id":"openai/gpt-4.1-mini-fast","name":"GPT-4.1 mini (Fast)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"cost":{"input":0.7,"output":2.8,"cache_read":0.175},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3 Pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-realtime-whisper":{"id":"openai/gpt-realtime-whisper","type":"transcription","name":"gpt-realtime-whisper","description":"Streaming speech-to-text model for low-latency transcript deltas from live audio","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/gpt-realtime-whisper"},"openai/gpt-image-1.5":{"id":"openai/gpt-image-1.5","type":"image","name":"GPT Image 1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":32,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1.5"},"openai/text-embedding-3-large":{"id":"openai/text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Most capable OpenAI embedding model for English and non-English text","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/text-embedding-3-large"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.625,"output":5,"cache_read":0.0625},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-4.1-fast":{"id":"openai/gpt-4.1-fast","name":"GPT-4.1 (Fast)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"cost":{"input":3.5,"output":14,"cache_read":0.875},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"flex":{"cost":{"input":0.025,"output":0.2,"cache_read":0.0025},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1-Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":1,"output":4,"cache_read":0.25},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT 5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":872000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/text-embedding-ada-002":{"id":"openai/text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Legacy OpenAI text embedding model","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/text-embedding-ada-002"},"openai/whisper-1":{"id":"openai/whisper-1","type":"transcription","name":"Whisper","description":"General-purpose speech recognition model for transcription and translation","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-03-01","last_updated":"2023-03-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/whisper-1"},"openai/gpt-live-1":{"id":"openai/gpt-live-1","type":"realtime","name":"GPT-Live 1","description":"Full-duplex voice model for natural conversations that delegates reasoning and tools to a backend agent","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/gpt-live-1"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT 5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.6-sol-fast":{"id":"openai/gpt-5.6-sol-fast","name":"GPT 5.6 Sol (Fast)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10,"tiers":[{"input":16,"output":60,"cache_read":1.6,"cache_write":20,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":16,"output":60,"cache_read":1.6,"cache_write":20}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/tts-1-hd":{"id":"openai/tts-1-hd","type":"speech","name":"TTS-1 HD","description":"Text-to-speech model optimized for audio quality","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/tts-1-hd"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5-codex"},"openai/gpt-5-fast":{"id":"openai/gpt-5-fast","name":"GPT-5 (Fast)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2.5,"output":20,"cache_read":0.25},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT 5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"amazon/nova-2-lite":{"id":"amazon/nova-2-lite","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.0825},"canonical_model_id":"amazon/nova-2-lite"},"amazon/titan-embed-text-v2":{"id":"amazon/titan-embed-text-v2","type":"embedding","name":"Titan Text Embeddings V2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"titan-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-04-30","last_updated":"2024-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"amazon/nova-lite":{"id":"amazon/nova-lite","name":"Nova Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015,"cache_write":0.06},"canonical_model_id":"amazon/nova-lite"},"amazon/nova-pro":{"id":"amazon/nova-pro","name":"Nova Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2,"cache_write":0.8},"canonical_model_id":"amazon/nova-pro"},"amazon/nova-micro":{"id":"amazon/nova-micro","name":"Nova Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875,"cache_write":0.035},"canonical_model_id":"amazon/nova-micro"},"perplexity/pplx-embed-v1-4b":{"id":"perplexity/pplx-embed-v1-4b","type":"embedding","name":"Embed v1 4b","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-02","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":8000},"cost":{"input":0.25,"output":2.5,"cache_write":0.0625}},"perplexity/pplx-embed-v1-0.6b":{"id":"perplexity/pplx-embed-v1-0.6b","type":"embedding","name":"Embed v1 0.6b","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"v0","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seed-1.8":{"id":"bytedance/seed-1.8","name":"Seed 1.8","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"cost":{"input":0.25,"output":2,"cache_read":0.05,"tiers":[{"input":0.5,"output":4,"cache_read":0.05,"tier":{"type":"context","size":128001}}]},"canonical_model_id":"bytedance-seed/seed-1-8"},"bytedance/seed-2.1-turbo":{"id":"bytedance/seed-2.1-turbo","name":"Seed 2.1 Turbo","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.5,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"bytedance/seedream-5.0-lite":{"id":"bytedance/seedream-5.0-lite","type":"image","name":"Seedream 5.0 Lite","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-13","last_updated":"2026-01-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedream-4.0":{"id":"bytedance/seedream-4.0","type":"image","name":"Seedream 4.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-09","last_updated":"2025-08-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-2.5":{"id":"bytedance/seedance-2.5","type":"video","name":"Seedance 2.5","description":"ByteDance multimodal audio-video model with up to 30-second single-pass generation, multi-round extension, reference inputs and timestamp-level editing.","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-31","last_updated":"2026-10-04","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"bytedance-seed/seedance-2.5"},"bytedance/seed-1.6":{"id":"bytedance/seed-1.6","name":"Seed 1.6","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.25,"output":2,"cache_read":0.05,"tiers":[{"input":0.5,"output":4,"cache_read":0.05,"tier":{"type":"context","size":128001}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"bytedance/seedance-2.0":{"id":"bytedance/seedance-2.0","type":"video","name":"Seedance 2.0","description":"ByteDance unified multimodal audio-video model supporting text, image, video and audio references, editing, extension and synchronized stereo sound.","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-12","last_updated":"2026-10-04","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"bytedance-seed/seedance-2.0"},"bytedance/seedream-5.0-pro":{"id":"bytedance/seedream-5.0-pro","type":"image","name":"Seedream 5.0 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-11","last_updated":"2026-07-11","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-v1.5-pro":{"id":"bytedance/seedance-v1.5-pro","type":"video","name":"Seedance v1.5 Pro","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-v1.0-pro":{"id":"bytedance/seedance-v1.0-pro","type":"video","name":"Seedance v1.0 Pro","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-11","last_updated":"2025-06-11","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-v1.0-pro-fast":{"id":"bytedance/seedance-v1.0-pro-fast","type":"video","name":"Seedance v1.0 Pro Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-24","last_updated":"2025-10-31","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedream-4.5":{"id":"bytedance/seedream-4.5","type":"image","name":"Seedream 4.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-11-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-2.0-fast":{"id":"bytedance/seedance-2.0-fast","type":"video","name":"Seedance 2.0 Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-14","last_updated":"2026-04-14","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-2.0-mini":{"id":"bytedance/seedance-2.0-mini","type":"video","name":"Seedance 2.0 Mini","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-06-22","last_updated":"2026-06-22","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/s1":{"id":"fish-audio/s1","type":"speech","name":"S1","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/drama-3":{"id":"fish-audio/drama-3","type":"speech","name":"Drama 3","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/s2.1-pro":{"id":"fish-audio/s2.1-pro","type":"speech","name":"S2.1 Pro","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/transcribe-1":{"id":"fish-audio/transcribe-1","type":"transcription","name":"Transcribe-1","description":"Speech transcription model for accurate audio-to-text and captioning workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-01","last_updated":"2026-03-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/s2-pro":{"id":"fish-audio/s2-pro","type":"speech","name":"S2 Pro","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k3.1-pro-highspeed":{"id":"moonshotai/kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4}},"moonshotai/kimi-k3-fast":{"id":"moonshotai/kimi-k3-fast","name":"Kimi K3 Fast","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code High Speed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"meta/llama-4-scout":{"id":"meta/llama-4-scout","name":"Llama-4-Scout-17B-16E-Instruct-FP8","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/llama-3.1-70b":{"id":"meta/llama-3.1-70b","name":"Llama 3.1 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-image-1.0":{"id":"meta/muse-image-1.0","type":"image","name":"Muse Image 1.0","description":"Agentic image generation and editing model that plans, searches and refines before rendering","family":"muse","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-07","last_updated":"2026-07-07","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"meta/muse-image-1.0"},"meta/llama-3.3-70b":{"id":"meta/llama-3.3-70b","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/llama-4-maverick":{"id":"meta/llama-4-maverick","name":"Llama-4-Maverick-17B-128E-Instruct-FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta/llama-3.1-8b":{"id":"meta/llama-3.1-8b","name":"Llama 3.1 8B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"bfl/flux-2-klein-9b":{"id":"bfl/flux-2-klein-9b","type":"image","name":"FLUX.2 [klein] 9B","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-3-video":{"id":"bfl/flux-3-video","type":"video","name":"Flux 3","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-04","last_updated":"2026-08-04","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-kontext-max":{"id":"bfl/flux-kontext-max","type":"image","name":"FLUX.1 Kontext Max","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-29","last_updated":"2025-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-pro-1.1-ultra":{"id":"bfl/flux-pro-1.1-ultra","type":"image","name":"FLUX1.1 [pro] Ultra","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-11-01","last_updated":"2024-11","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-max":{"id":"bfl/flux-2-max","type":"image","name":"FLUX.2 [max]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":67300,"output":67300},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-pro":{"id":"bfl/flux-2-pro","type":"image","name":"FLUX.2 [pro]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":67300,"output":67300},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-3-image-klein":{"id":"bfl/flux-3-image-klein","type":"image","name":"FLUX 3 Image Klein","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-kontext-pro":{"id":"bfl/flux-kontext-pro","type":"image","name":"FLUX.1 Kontext Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-29","last_updated":"2025-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-flex":{"id":"bfl/flux-2-flex","type":"image","name":"FLUX.2 [flex]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-pro-1.1":{"id":"bfl/flux-pro-1.1","type":"image","name":"FLUX1.1 [pro]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-02","last_updated":"2024-10","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-pro-1.0-fill":{"id":"bfl/flux-pro-1.0-fill","type":"image","name":"FLUX.1 Fill [pro]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-01","last_updated":"2024-10","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-klein-4b":{"id":"bfl/flux-2-klein-4b","type":"image","name":"FLUX.2 [klein] 4B","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-3-image":{"id":"bfl/flux-3-image","type":"image","name":"FLUX 3 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"elorian/glimpse":{"id":"elorian/glimpse","name":"Glimpse","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":220000,"output":4096},"cost":{"input":0.6,"output":3,"cache_read":0.06}},"liquid/lfm2.5-24b-a2b":{"id":"liquid/lfm2.5-24b-a2b","name":"LFM2.5-24B-A2B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8200},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"stealth/glyph-cluster":{"id":"stealth/glyph-cluster","name":"Glyph Cluster","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0,"output":0}},"callstack/apex":{"id":"callstack/apex","name":"Apex","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"release_date":"2026-10-08","last_updated":"2026-10-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.2}},"tencent/hy-mt2-pro":{"id":"tencent/hy-mt2-pro","name":"Tencent Hy-MT2-Pro","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":4000},"cost":{"input":0.074,"output":0.295}},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"tencent/hy-mt2-plus":{"id":"tencent/hy-mt2-plus","name":"Tencent Hy-MT2-Plus","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":4000},"cost":{"input":0.074,"output":0.295}},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Tencent Hy4 Preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy-mt2-lite":{"id":"tencent/hy-mt2-lite","name":"Tencent Hy-MT2-Lite","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":4000},"cost":{"input":0.044,"output":0.177}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.05},"canonical_model_id":"inference-net/schematron-v2-small"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100001}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4.8-fast":{"id":"anthropic/claude-opus-4.8-fast","name":"Claude Opus 4.8 (Fast)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-5-fast":{"id":"anthropic/claude-opus-5-fast","name":"Claude Opus 5 (Fast)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-5.5-fast":{"id":"anthropic/claude-opus-5.5-fast","name":"Claude Opus 5.5 (Fast)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":8,"output":40,"cache_read":0.4,"cache_write":10},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"interfaze/interfaze-beta":{"id":"interfaze/interfaze-beta","name":"Interfaze Beta","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"temperature":true,"release_date":"2025-10-07","last_updated":"2026-04-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":1.5,"output":3.5}},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"NVIDIA Nemotron 3.5 Super VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.2,"output":0.5,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3.5-lightning-free":{"id":"nvidia/nemotron-3.5-lightning-free","name":"Nemotron 3.5 Lightning Free","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nemotron 3.5 Lightning 30B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"NVIDIA Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.15,"output":0.65},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-nano-9b-v2":{"id":"nvidia/nemotron-nano-9b-v2","name":"Nvidia Nemotron Nano 9B V2","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-08-18","last_updated":"2025-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.06,"output":0.23},"canonical_model_id":"nvidia/nemotron-nano-9b-v2"},"nvidia/nemotron-nano-12b-v2-vl":{"id":"nvidia/nemotron-nano-12b-v2-vl","name":"Nvidia Nemotron Nano 12B V2 VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.2,"output":0.6},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"prodia/flux-fast-schnell":{"id":"prodia/flux-fast-schnell","type":"image","name":"Flux Schnell","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-02","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"meituan/longcat-2.5-preview":{"id":"meituan/longcat-2.5-preview","name":"LongCat 2.5 Preview","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.5-preview"},"mixedbread/toast-1":{"id":"mixedbread/toast-1","name":"Toast 1","description":"Specialized search model for knowledge-intensive questions, multi-step retrieval, and evidence synthesis","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131000,"output":4000},"cost":{"input":0.3,"output":0.72,"cache_read":0.036},"canonical_model_id":"mixedbread/toast-1"},"klingai/kling-v2.5-turbo-i2v":{"id":"klingai/kling-v2.5-turbo-i2v","type":"video","name":"Kling v2.5 Turbo Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.6-motion-control":{"id":"klingai/kling-v2.6-motion-control","type":"video","name":"Kling v2.6 Motion Control","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-21","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v3.0-i2v":{"id":"klingai/kling-v3.0-i2v","type":"video","name":"Kling v3.0 Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.6-i2v":{"id":"klingai/kling-v2.6-i2v","type":"video","name":"Kling v2.6 Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-12-21","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.5-turbo-t2v":{"id":"klingai/kling-v2.5-turbo-t2v","type":"video","name":"Kling v2.5 Turbo Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v3.0-t2v":{"id":"klingai/kling-v3.0-t2v","type":"video","name":"Kling v3.0 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v3.0-motion-control":{"id":"klingai/kling-v3.0-motion-control","type":"video","name":"Kling v3.0 Motion Control","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-04","last_updated":"2026-03-04","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.6-t2v":{"id":"klingai/kling-v2.6-t2v","type":"video","name":"Kling v2.6 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember-1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"}}},"neon":{"id":"neon","env":["NEON_AI_GATEWAY_BASE_URL","NEON_AI_GATEWAY_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"${NEON_AI_GATEWAY_BASE_URL}/v1","name":"Neon","doc":"https://neon.com/docs","models":{"gemini-3-7-flash":{"id":"gemini-3-7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.7-flash"},"meta-llama-3-1-8b-instruct":{"id":"meta-llama-3-1-8b-instruct","name":"Llama 3.1 8B Instruct","description":"Meta's compact open-weight Llama 3.1 model for fast, low-cost text generation","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.45},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"glm-5-3-flash":{"id":"glm-5-3-flash","name":"GLM-5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"meta-llama-3-3-70b-instruct":{"id":"meta-llama-3-3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gemini-3-1-pro":{"id":"gemini-3-1-pro","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"glm-5-2":{"id":"glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"gemini-3-5-flash-lite":{"id":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5-5-pro":{"id":"gpt-5-5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"gemini-3-flash":{"id":"gemini-3-flash","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"qwen35-122b-a10b":{"id":"qwen35-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":25000},"cost":{"input":0.22,"output":2.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"gemini-3-6-flash":{"id":"gemini-3-6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-5-4-mini":{"id":"gpt-5-4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-5-4":{"id":"gpt-5-4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gpt-5-1":{"id":"gpt-5-1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"gpt-5-5":{"id":"gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5-6-terra":{"id":"gpt-5-6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"grok-4-7":{"id":"grok-4-7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"gemma-3-12b":{"id":"gemma-3-12b","name":"Gemma 3 12B","description":"Google's open-weight Gemma 3 vision-language model for text and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.5},"canonical_model_id":"google/gemma-3-12b-it"},"gemini-3-8-flash":{"id":"gemini-3-8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.8-flash"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-5-2":{"id":"gpt-5-2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"llama-4-maverick":{"id":"llama-4-maverick","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-6-1-sol":{"id":"gpt-6-1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"gemini-3-1-flash-lite":{"id":"gemini-3-1-flash-lite","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"gpt-5-6-luna":{"id":"gpt-5-6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.05,"output":0.4,"cache_read":0.01},"canonical_model_id":"openai/gpt-5-nano"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":524288},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gpt-5-4-nano":{"id":"gpt-5-4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":10000},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"gpt-5-3-codex":{"id":"gpt-5-3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"glm-5-3":{"id":"glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"gemini-3-5-flash":{"id":"gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5-6-sol":{"id":"gpt-5-6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"}}},"kosmik":{"id":"kosmik","env":["KOSMIK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.koscompute.com/v1","name":"Kosmik Compute","doc":"https://api.koscompute.com/docs/","models":{"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.35,"output":2.2,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"volcengine":{"id":"volcengine","env":["ARK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ark.cn-beijing.volces.com/api/v3","name":"Volcengine Ark","doc":"https://www.volcengine.com/docs/82379/1330310","models":{"doubao-seed-2-0-mini-260428":{"id":"doubao-seed-2-0-mini-260428","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.02969,"output":0.29687,"cache_read":0.00594,"tiers":[{"input":0.05937,"output":0.59374,"cache_read":0.01187,"tier":{"type":"context","size":32000}},{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"glm-5-2-260617":{"id":"glm-5-2-260617","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.18747,"output":4.15615,"cache_read":0.29687},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4-flash-ga-260731":{"id":"deepseek-v4-flash-ga-260731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.4453,"output":1.3359,"cache_read":0.01484},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-v4-pro-ga-260813":{"id":"deepseek-v4-pro-ga-260813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.3359,"output":4.00771,"cache_read":0.04453},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"glm-5-3-flash-260828":{"id":"glm-5-3-flash-260828","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11875,"output":0.41563,"cache_read":0.03414},"canonical_model_id":"zhipuai/glm-5.3-flash"},"doubao-seed-2-0-pro-260215":{"id":"doubao-seed-2-0-pro-260215","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.47499,"output":2.37494,"cache_read":0.095,"tiers":[{"input":0.71248,"output":3.56241,"cache_read":0.1425,"tier":{"type":"context","size":32000}},{"input":1.42496,"output":7.12482,"cache_read":0.28499,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"doubao-seed-1-6-vision-250815":{"id":"doubao-seed-1-6-vision-250815","name":"Seed 1.6 Vision","description":"ByteDance Seed multimodal model for image understanding, visual reasoning, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-08-15","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":2.37494,"cache_read":0.02375,"tier":{"type":"context","size":32000}},{"input":0.35624,"output":3.56241,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-vision"},"doubao-seed-1-8-251228":{"id":"doubao-seed-1-8-251228","name":"Seed 1.8","description":"ByteDance Seed model for multimodal reasoning, long-context analysis, and agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":2.37494,"cache_read":0.02375,"tier":{"type":"context","size":32000}},{"input":0.35624,"output":3.56241,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-8"},"doubao-seed-2-1-pro-260628":{"id":"doubao-seed-2-1-pro-260628","name":"Seed 2.1 Pro","description":"Flagship ByteDance Seed 2.1 model for complex multimodal reasoning, coding, and agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.8906,"output":4.45301,"cache_read":0.17812},"canonical_model_id":"bytedance-seed/seed-2.1-pro"},"doubao-seed-1-6-flash-250828":{"id":"doubao-seed-1-6-flash-250828","name":"Seed 1.6 Flash","description":"Low-latency ByteDance Seed model for high-throughput chat, extraction, and lightweight tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.02227,"output":0.22265,"cache_read":0.00445,"tiers":[{"input":0.04453,"output":0.4453,"cache_read":0.00445,"tier":{"type":"context","size":32000}},{"input":0.08906,"output":0.8906,"cache_read":0.00445,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"doubao-seed-character-260628":{"id":"doubao-seed-character-260628","name":"Seed Character","description":"ByteDance Seed model optimized for character-driven dialogue and consistent conversational behavior","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.11875,"output":0.29687,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":0.8906,"cache_read":0.02375,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"bytedance-seed/seed-character"},"doubao-seed-2-0-code-preview-260215":{"id":"doubao-seed-2-0-code-preview-260215","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.47499,"output":2.37494,"cache_read":0.095,"tiers":[{"input":0.71248,"output":3.56241,"cache_read":0.1425,"tier":{"type":"context","size":32000}},{"input":1.42496,"output":7.12482,"cache_read":0.28499,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"doubao-seed-2-1-turbo-260628":{"id":"doubao-seed-2-1-turbo-260628","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.4453,"output":2.22651,"cache_read":0.08906},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"doubao-seed-evolving":{"id":"doubao-seed-evolving","name":"Seed Evolving","description":"Rolling ByteDance Seed model for rapidly updated reasoning, coding, and agent capabilities","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.8906,"output":4.45301,"cache_read":0.17812},"canonical_model_id":"bytedance-seed/seed-evolving"},"doubao-seed-1-6-251015":{"id":"doubao-seed-1-6-251015","name":"Seed 1.6","description":"ByteDance Seed model for long-context reasoning, instruction following, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":2.37494,"cache_read":0.02375,"tier":{"type":"context","size":32000}},{"input":0.35624,"output":3.56241,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"doubao-seed-2-0-lite-260428":{"id":"doubao-seed-2-0-lite-260428","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.08906,"output":0.53436,"cache_read":0.01781,"tiers":[{"input":0.13359,"output":0.80154,"cache_read":0.02672,"tier":{"type":"context","size":32000}},{"input":0.26718,"output":1.60308,"cache_read":0.05344,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"}}},"wandb":{"id":"wandb","env":["WANDB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inference.wandb.ai/v1","name":"CoreWeave","doc":"https://docs.wandb.ai/inference","models":{"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B A3B","description":"Qwen3.6-35B-A3B is an MoE multimodal model with 262K context optimized for agentic coding workflows.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":1.25,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Qwen3.8-27B is a dense multimodal model suited for coding, research, vision, and long-running agent tasks.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":3,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.8-27b"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax M3","description":"MiniMax M3 is a multimodal MoE model with 23B active parameters optimized for coding and agentic workflows.","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.23,"output":0.96,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"reflection/Beam-1.0-501B-A23B":{"id":"reflection/Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B","description":"Beam is a 501B-parameter MoE model with 23B active parameters built for coding, reasoning, and agentic workflows.","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1,"cache_read":0.04},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B","description":"Granite 4.2 8B is an instruct model capable of enhanced tool calling, instruction following, and chat capabilities.","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.15,"cache_read":0.05},"canonical_model_id":"ibm/granite-4.2-8b"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B","description":"Gemma 4 31B Dense is designed for advanced reasoning, agentic workflows, and longer context and is natively trained on 140+ languages.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.34,"cache_read":0.1},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B","description":"Gemma 4 26B A4B is a multimodal MoE model with LoRA support and function calling for agentic workflows.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"gpt-oss-20b","description":"Lower latency Mixture-of-Experts model trained on OpenAI's Harmony response format with reasoning capabilities.","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.03,"output":0.13,"cache_read":0.03},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"gpt-oss-120b","description":"Efficient Mixture-of-Experts model designed for high-reasoning, agentic and general-purpose use cases.","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.03,"output":0.17,"cache_read":0.03},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"DeepSeek V4-Flash-0731 is an MoE model great for coding, reasoning, and agentic workloads.","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.13,"output":0.28,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4-Pro-0813 is a 1.6T-parameter MoE model excelling at advanced reasoning, coding, and complex agentic workloads.","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.31,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"A large hybrid model that supports both thinking and non-thinking modes via prompt templates.","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":161000,"output":161000},"cost":{"input":0.55,"output":1.65,"cache_read":0.55},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash is a multimodal MoE model for coding, reasoning, and agentic workloads with long contexts.","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.65,"cache_read":0.002},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Kimi K2.6 is a multimodal Mixture-of-Experts language model featuring 32 billion activated parameters and a total of 1 trillion parameters.","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.65,"output":3.41,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Kimi K2.7 Code is a 1T-parameter MoE model with 32B active parameters purpose-built for long-horizon agentic coding and software engineering.","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.71,"output":3.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"meta-llama/Llama-3.1-8B-Instruct":{"id":"meta-llama/Llama-3.1-8B-Instruct","name":"Llama 3.1 8B","description":"Efficient conversational model optimized for responsive multilingual chatbot interactions.","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.22,"output":0.22,"cache_read":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B","description":"Multilingual model excelling in conversational tasks, detailed instruction-following, and coding.","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.71,"output":0.71,"cache_read":0.71},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B","name":"Nemotron 3 Ultra","description":"Nemotron 3 Ultra is a powerful MoE model designed for long-running agents across coding, deep research, and enterprise automation.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.15,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B","name":"Nemotron 3.5 Super VL","description":"Nemotron 3.5 Super VL is a multimodal MoE model built for visual understanding, reasoning, and agentic workflows.","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B","name":"Nemotron 3.5 Lightning","description":"Nemotron 3.5 Lightning is an MoE model built for fast, reliable agentic tasks across use cases such as financial services, cybersecurity, telecom, and retail.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.07,"output":0.2,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM 5.3","description":"GLM-5.3 is a 744B-parameter MoE model built for advanced reasoning, software engineering, and long-horizon agentic tasks.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.2,"output":4,"cache_read":0.19},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"GLM-5.3-Flash is a natively multimodal model with 320B total parameters and 18B active parameters.","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM 5.2","description":"GLM-5.2 is a Mixture-of-Experts language model featuring 40 billion activated parameters and a total of 744 billion parameters.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.62,"output":2.42,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"}}},"alibaba-cn":{"id":"alibaba-cn","env":["DASHSCOPE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://dashscope.aliyuncs.com/compatible-mode/v1","name":"Alibaba (China)","doc":"https://www.alibabacloud.com/help/en/model-studio/models","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-11","last_updated":"2026-02-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0.573,"output":2.58,"tiers":[{"input":0.86,"output":3.154,"tier":{"type":"context","size":32000}}]}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.29754,"output":1.19015,"cache_read":0.01488},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"qwen-vl-max":{"id":"qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.23,"output":0.574}},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1,"output":3.851,"cache_read":0.275,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-v3-1":{"id":"deepseek-v3-1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0.574,"output":1.721}},"deepseek-r1-distill-qwen-32b":{"id":"deepseek-r1-distill-qwen-32b","name":"DeepSeek R1 Distill Qwen 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.287,"output":0.861}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":2.827,"output":14.133,"cache_read":0.283},"canonical_model_id":"moonshotai/kimi-k3"},"qwen2-5-omni-7b":{"id":"qwen2-5-omni-7b","name":"Qwen2.5-Omni 7B","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-12","last_updated":"2024-12","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":true,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.087,"output":0.345,"input_audio":5.448}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.276,"output":1.651,"cache_read":0.0276,"cache_write":0.345,"tiers":[{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625}}},"tongyi-intent-detect-v3":{"id":"tongyi-intent-detect-v3","name":"Tongyi Intent Detect V3","description":"General-purpose chat model for instruction following, writing, and analysis","family":"yi","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1024},"cost":{"input":0.058,"output":0.144}},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen3 235B-A22B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.287,"output":1.147,"reasoning":2.868}},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.65,"output":3.301},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen-mt-plus":{"id":"qwen-mt-plus","name":"Qwen-MT Plus","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":0.259,"output":0.775}},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.216,"output":0.861,"tiers":[{"input":0.323,"output":1.291,"tier":{"type":"context","size":32000}},{"input":0.538,"output":2.151,"tier":{"type":"context","size":128000}}]}},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-14","last_updated":"2026-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0.825,"output":3.301,"cache_read":0.165,"cache_write":1.03125,"tiers":[{"input":1.1,"output":3.851,"cache_read":0.22,"cache_write":1.375,"tier":{"type":"context","size":32000}}]}},"deepseek-r1-distill-qwen-7b":{"id":"deepseek-r1-distill-qwen-7b","name":"DeepSeek R1 Distill Qwen 7B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.072,"output":0.144}},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.287,"output":1.147,"reasoning":2.868}},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":131072},"cost":{"input":0.02962,"output":0.1185,"cache_read":0.005924,"cache_write":0.03703,"tiers":[{"input":0.08887,"output":0.35549,"cache_read":0.017774,"cache_write":0.11109,"tier":{"type":"context","size":32000}},{"input":0.17774,"output":0.71098,"cache_read":0.035548,"cache_write":0.22218,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen-omni-turbo-realtime":{"id":"qwen-omni-turbo-realtime","type":"realtime","name":"Qwen-Omni Turbo Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-05-08","last_updated":"2025-05-08","modalities":{"input":["text","image","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","provider":{"api":"wss://dashscope.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.23,"output":0.918,"input_audio":3.584,"output_audio":7.168}},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"Qwen3-VL 235B-A22B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.286705,"output":1.14682,"reasoning":2.867051}},"qwen3-coder-flash":{"id":"qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.144,"output":0.574,"cache_read":0.0288,"cache_write":0.18,"tiers":[{"input":0.216,"output":0.861,"cache_read":0.0432,"cache_write":0.27,"tier":{"type":"context","size":32000}},{"input":0.359,"output":1.434,"cache_read":0.0718,"cache_write":0.44875,"tier":{"type":"context","size":128000}},{"input":0.717,"output":3.584,"cache_read":0.1434,"cache_write":0.89625,"tier":{"type":"context","size":256000}}]}},"qwen-math-plus":{"id":"qwen-math-plus","name":"Qwen Math Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-08-16","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":3072},"status":"deprecated","cost":{"input":0.574,"output":1.721}},"deepseek-r1-distill-qwen-14b":{"id":"deepseek-r1-distill-qwen-14b","name":"DeepSeek R1 Distill Qwen 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.144,"output":0.431}},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.65,"output":4.951,"cache_read":0.33,"cache_write":2.0625},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen3-next-80b-a3b-thinking":{"id":"qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.144,"output":1.434}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Moonshot Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.574,"output":3.011,"cache_read":0.1148,"cache_write":0.7175}},"moonshot-kimi-k2-instruct":{"id":"moonshot-kimi-k2-instruct","name":"Moonshot Kimi K2 Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"qwen3-8b":{"id":"qwen3-8b","name":"Qwen3 8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.072,"output":0.287,"reasoning":0.717}},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.143353,"output":1.433525,"cache_read":0.028671,"cache_write":0.179191,"tiers":[{"input":0.215,"output":2.15,"cache_read":0.043,"cache_write":0.26875,"tier":{"type":"context","size":32000}},{"input":0.43,"output":4.301,"cache_read":0.086,"cache_write":0.5375,"tier":{"type":"context","size":128000}}]}},"qwen-plus-character":{"id":"qwen-plus-character","name":"Qwen Plus Character","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"cost":{"input":0.115,"output":0.287}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2026-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.574,"output":2.296,"tiers":[{"input":0.861,"output":3.444,"tier":{"type":"context","size":32000}},{"input":1.435,"output":5.74,"tier":{"type":"context","size":128000}},{"input":2.87,"output":28.7,"tier":{"type":"context","size":256000}}]}},"deepseek-r1-distill-qwen-1-5b":{"id":"deepseek-r1-distill-qwen-1-5b","name":"DeepSeek R1 Distill Qwen 1.5B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"cost":{"input":0,"output":0}},"qwen-vl-ocr":{"id":"qwen-vl-ocr","name":"Qwen-VL OCR","description":"OCR model for extracting structured text from documents and screenshots","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-10-28","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":34096,"output":4096},"cost":{"input":0.043,"output":0.072}},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-23","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.359,"output":1.434,"reasoning":1.434,"tiers":[{"input":0.574,"output":2.294,"reasoning":2.294,"tier":{"type":"context","size":32000}},{"input":1.004,"output":4.014,"reasoning":4.014,"tier":{"type":"context","size":128000}}]}},"qwen-flash":{"id":"qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.022,"output":0.216,"cache_read":0.0044,"cache_write":0.0275,"tiers":[{"input":0.087,"output":0.861,"cache_read":0.0174,"cache_write":0.10875,"tier":{"type":"context","size":128000}},{"input":0.173,"output":1.721,"cache_read":0.0346,"cache_write":0.21625,"tier":{"type":"context","size":256000}}]}},"qwen3.8-omni-flash":{"id":"qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.113,"output":0.382,"cache_read":0.014},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"qwen-long":{"id":"qwen-long","name":"Qwen Long","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-25","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"output":8192},"cost":{"input":0.072,"output":0.287}},"qwen3-omni-flash-realtime":{"id":"qwen3-omni-flash-realtime","type":"realtime","name":"Qwen3-Omni Flash Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"status":"deprecated","provider":{"api":"wss://dashscope.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.23,"output":0.918,"input_audio":3.584,"output_audio":7.168}},"qwen3.6-max-preview":{"id":"qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":245800,"output":65536},"status":"deprecated","cost":{"input":1.32,"output":7.9,"cache_read":0.132}},"qwen-turbo":{"id":"qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":16384},"status":"deprecated","cost":{"input":0.044,"output":0.087,"reasoning":0.431}},"qwen-deep-research":{"id":"qwen-deep-research","name":"Qwen Deep Research","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":7.742,"output":23.367}},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.172,"output":1.032,"reasoning":1.032,"tiers":[{"input":0.43,"output":2.58,"reasoning":2.58,"tier":{"type":"context","size":128000}}]}},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.77744,"output":5.33231,"cache_read":0.22218,"cache_write":2.22179},"canonical_model_id":"alibaba/qwen3.8-max"},"deepseek-v3":{"id":"deepseek-v3","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-12-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"status":"deprecated","cost":{"input":0.287,"output":1.147}},"qwen-math-turbo":{"id":"qwen-math-turbo","name":"Qwen Math Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":3072},"status":"deprecated","cost":{"input":0.287,"output":0.861}},"kimi-k2.6":{"id":"kimi-k2.6","name":"Moonshot Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.929,"output":3.858,"cache_read":0.1858,"cache_write":1.16125}},"qvq-max":{"id":"qvq-max","name":"QVQ Max","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qvq","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":1.147,"output":4.588}},"qwen-mt-turbo":{"id":"qwen-mt-turbo","name":"Qwen-MT Turbo","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"status":"deprecated","cost":{"input":0.101,"output":0.28}},"qwen3-14b":{"id":"qwen3-14b","name":"Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.144,"output":0.574,"reasoning":1.434}},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-09-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.115,"output":0.688,"reasoning":0.688,"tiers":[{"input":0.287,"output":1.72,"reasoning":1.72,"tier":{"type":"context","size":128000}},{"input":0.573,"output":3.44,"reasoning":3.44,"tier":{"type":"context","size":256000}}]}},"qwen3-vl-30b-a3b":{"id":"qwen3-vl-30b-a3b","name":"Qwen3-VL 30B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.108,"output":0.431,"reasoning":1.076}},"deepseek-r1-0528":{"id":"deepseek-r1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.1,"output":3.851,"cache_read":0.275,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2}},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11875,"output":0.40073,"cache_read":0.014844,"cache_write":0.185547},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3-asr-flash":{"id":"qwen3-asr-flash","type":"transcription","name":"Qwen3-ASR Flash","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-04","release_date":"2025-09-08","last_updated":"2025-09-08","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":53248,"output":4096}},"qwen-plus":{"id":"qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.115,"output":0.287,"reasoning":1.147,"cache_read":0.023,"cache_write":0.144,"tiers":[{"input":0.345,"output":2.868,"reasoning":3.441,"cache_read":0.069,"cache_write":0.43125,"tier":{"type":"context","size":128000}},{"input":0.689,"output":6.881,"reasoning":9.175,"cache_read":0.1378,"cache_write":0.86125,"tier":{"type":"context","size":256000}}]}},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.276,"output":1.101,"cache_read":0.0552,"cache_write":0.345,"tiers":[{"input":0.826,"output":3.301,"cache_read":0.1652,"cache_write":1.0325,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.826,"output":3.301,"cache_read":0.1652,"cache_write":1.0325}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen-max":{"id":"qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.345,"output":1.377}},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Moonshot Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"qwen-vl-plus":{"id":"qwen-vl-plus","name":"Qwen-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.115,"output":0.287}},"qwen-omni-turbo":{"id":"qwen-omni-turbo","name":"Qwen-Omni Turbo","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-19","last_updated":"2025-03-26","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.058,"output":0.23,"input_audio":3.584,"output_audio":7.168}},"qwen3.5-flash":{"id":"qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-09-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.029,"output":0.287,"reasoning":0.287,"tiers":[{"input":0.115,"output":1.147,"reasoning":1.147,"tier":{"type":"context","size":128000}},{"input":0.172,"output":1.72,"reasoning":1.72,"tier":{"type":"context","size":256000}}]}},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.861,"output":3.441,"tiers":[{"input":1.291,"output":5.161,"tier":{"type":"context","size":32000}},{"input":2.151,"output":8.602,"tier":{"type":"context","size":128000}}]}},"qwen-doc-turbo":{"id":"qwen-doc-turbo","name":"Qwen Doc Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.087,"output":0.144}},"deepseek-r1-distill-llama-70b":{"id":"deepseek-r1-distill-llama-70b","name":"DeepSeek R1 Distill Llama 70B","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"cost":{"input":0.287,"output":0.861}},"deepseek-v3-2-exp":{"id":"deepseek-v3-2-exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0.287,"output":0.431}},"qwen3-omni-flash":{"id":"qwen3-omni-flash","name":"Qwen3-Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"cost":{"input":0.258,"output":0.989,"input_audio":2.265,"output_audio":8.974}},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.144,"output":0.574}},"deepseek-r1":{"id":"deepseek-r1","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"qwq-plus":{"id":"qwq-plus","name":"QwQ Plus","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.23,"output":0.574}},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.165,"output":0.99,"cache_read":0.0165,"cache_write":0.20625,"tiers":[{"input":0.66,"output":3.961,"cache_read":0.066,"cache_write":0.825,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.66,"output":3.961,"cache_read":0.066,"cache_write":0.825}},"canonical_model_id":"alibaba/qwen3.6-flash"},"deepseek-r1-distill-llama-8b":{"id":"deepseek-r1-distill-llama-8b","name":"DeepSeek R1 Distill Llama 8B","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0,"output":0}},"kimi/kimi-k2.5":{"id":"kimi/kimi-k2.5","name":"kimi/kimi-k2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1}},"MiniMax/MiniMax-M2.7":{"id":"MiniMax/MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375}},"siliconflow/deepseek-v3-0324":{"id":"siliconflow/deepseek-v3-0324","name":"siliconflow/deepseek-v3-0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":163840},"cost":{"input":0.25,"output":1}},"siliconflow/deepseek-v3.1-terminus":{"id":"siliconflow/deepseek-v3.1-terminus","name":"siliconflow/deepseek-v3.1-terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":65536},"cost":{"input":0.27,"output":1}},"siliconflow/deepseek-r1-0528":{"id":"siliconflow/deepseek-r1-0528","name":"siliconflow/deepseek-r1-0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.18}},"siliconflow/deepseek-v3.2":{"id":"siliconflow/deepseek-v3.2","name":"siliconflow/deepseek-v3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-03","last_updated":"2025-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":65536},"cost":{"input":0.27,"output":0.42}}}},"vancine":{"id":"vancine","env":["VANCINE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://vancine.com/v1","name":"Vancine","doc":"https://vancine.com/docs","models":{"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.24,"output":0.96,"cache_read":0.0048},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.12,"output":3.52,"cache_read":0.208},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.4,"output":12,"cache_read":0.24},"canonical_model_id":"moonshotai/kimi-k3"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"status":"deprecated","cost":{"input":0.24,"output":0.96,"cache_read":0.048},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.6,"output":4.8,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.12,"output":0.38,"cache_read":0.013},"canonical_model_id":"alibaba/qwen3.8-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.12,"output":0.4,"cache_read":0.024},"canonical_model_id":"zhipuai/glm-5.3-flash"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.67,"output":2,"cache_read":0.034},"canonical_model_id":"tencent/hy4-preview"}}},"poolside":{"id":"poolside","env":["POOLSIDE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.poolside.ai/v1","name":"Poolside","doc":"https://platform.poolside.ai","models":{"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-m.1":{"id":"poolside/laguna-m.1","name":"Laguna M.1","description":"Poolside's open-weight model for agentic coding and long-horizon work","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"poolside/laguna-m.1"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"poolside/laguna-xs-2.1"}}},"hpc-ai":{"id":"hpc-ai","env":["HPC_AI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.hpc-ai.com/inference/v1","name":"HPC-AI","doc":"https://www.hpc-ai.com/doc/docs/quickstart/","models":{"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1002000,"output":128000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"output":195000},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-7"},"zai-org/glm-5.1":{"id":"zai-org/glm-5.1","name":"GLM 5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202000,"output":202000},"cost":{"input":0.615,"output":2.46,"cache_read":0.133},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"above":{"id":"above","env":["ABOVE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.above.dev/v1","name":"above.dev","doc":"https://above.dev/docs","models":{"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.165,"output":0.66,"reasoning":0.66,"cache_read":0.0033},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.726,"output":2.178,"reasoning":2.178,"cache_read":0.0242},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1692,"output":0.3385,"cache_read":0.0034},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"glm-5.2-fast":{"id":"glm-5.2-fast","name":"GLM 5.2 Fast","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.31,"output":7.26,"cache_read":0.231},"canonical_model_id":"zhipuai/glm-5.2"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.5077,"output":1.0154,"cache_read":0.0042},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.165,"output":0.66,"reasoning":0.66,"cache_read":0.0033},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"mimo-v2.6-pro-ultraspeed":{"id":"mimo-v2.6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":5.0769,"output":10.1538,"cache_read":0.0423},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen 3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2.2,"output":6.6,"cache_read":0.275},"canonical_model_id":"alibaba/qwen3.8-max"},"glm-5.2":{"id":"glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.54,"output":4.84,"cache_read":0.154},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.165,"output":0.55,"cache_read":0.0319},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"vivgrid":{"id":"vivgrid","env":["VIVGRID_API_KEY"],"npm":"@ai-sdk/openai","api":"https://api.vivgrid.com/v1","name":"Vivgrid","doc":"https://docs.vivgrid.com/models","models":{"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.31,"output":1.23,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.2,"output":4.2,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.3"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT 5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-terra"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.74,"output":3.48,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT 5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":1,"output":6,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.44,"output":1.32,"reasoning":1.32,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":1.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.35,"output":3,"reasoning":3,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":0.5,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":3,"output":12,"cache_read":0.3,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.2,"output":4.2,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.15,"output":0.5,"cache_read":0.04},"canonical_model_id":"zhipuai/glm-5.3-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.75,"output":3.75,"cache_read":0.15},"canonical_model_id":"google/gemini-3.8-flash"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT 5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.28,"output":0.42,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v3.2"},"viv-fast":{"id":"viv-fast","name":"Viv Fast","description":"Fast coding model","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-09","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":256000},"cost":{"input":0.13,"output":0.4,"cache_read":0.05},"canonical_model_id":"vivgrid/viv-fast"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"modal":{"id":"modal","env":["MODAL_PROXY_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.us-west.modal.direct/v1","name":"Modal","doc":"https://modal.com/docs/guide/endpoints","models":{"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8-Max","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"thinkingmachines/Inkling-NVFP4":{"id":"thinkingmachines/Inkling-NVFP4","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.2,"output":5,"cache_read":0.27},"canonical_model_id":"thinkingmachines/inkling"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"meta-models/Muse-Spark-Open-816B-A42B":{"id":"meta-models/Muse-Spark-Open-816B-A42B","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"reasoning":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"nvidia/GLM-5.3-Flash-NVFP4":{"id":"nvidia/GLM-5.3-Flash-NVFP4","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.45,"output":1.5,"cache_read":0.09},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"claudinio":{"id":"claudinio","env":["CLAUDINIO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.claudin.io/v1","name":"Claudinio","doc":"https://claudin.io","models":{"claudius":{"id":"claudius","name":"Claudius","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"knowledge":"2026-05","release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":3,"output":8,"cache_read":0.9}},"claudinio":{"id":"claudinio","name":"Claudinio","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"knowledge":"2026-05","release_date":"2026-05-12","last_updated":"2026-06-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":0.5,"output":2,"cache_read":0.15}}}},"stepfun-ai-step-plan":{"id":"stepfun-ai-step-plan","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.ai/step_plan/v1","name":"StepFun Step Plan (Global)","doc":"https://platform.stepfun.ai/docs/en/step-plan/integrations/reasoning-api","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.7-flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash"}}},"atomic-chat":{"id":"atomic-chat","env":["ATOMIC_CHAT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://127.0.0.1:1337/v1","name":"Atomic Chat","doc":"https://atomic.chat","models":{"gemma-4-E4B-it-MLX-4bit":{"id":"gemma-4-E4B-it-MLX-4bit","name":"Gemma 4 E4B Instruct (MLX 4-bit)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-E4B-it"},"gemma-4-E4B-it-IQ4_XS":{"id":"gemma-4-E4B-it-IQ4_XS","name":"Gemma 4 E4B Instruct (IQ4_XS)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-E4B-it"},"Qwen3_5-9B-Q4_K_M":{"id":"Qwen3_5-9B-Q4_K_M","name":"Qwen 3.5 9B (Q4_K_M)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen3_5-9B-MLX-4bit":{"id":"Qwen3_5-9B-MLX-4bit","name":"Qwen 3.5 9B (MLX 4-bit)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-9b"},"Meta-Llama-3_1-8B-Instruct-GGUF":{"id":"Meta-Llama-3_1-8B-Instruct-GGUF","name":"Meta Llama 3.1 8B Instruct (GGUF)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.1-8b-instruct"}}},"xiaomi":{"id":"xiaomi","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.xiaomimimo.com/v1","name":"Xiaomi","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"MiMo Pro model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo-v2.6-pro-ultraspeed":{"id":"mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5-pro-ultraspeed":{"id":"mimo-v2.5-pro-ultraspeed","name":"MiMo-V2.5-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-06-08","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.305,"output":2.61,"cache_read":0.0108},"canonical_model_id":"xiaomi/mimo-v2.5-pro-ultraspeed"}}},"sap-ai-core":{"id":"sap-ai-core","env":["AICORE_SERVICE_KEY"],"npm":"@jerome-benoit/sap-ai-provider-v2","name":"SAP AI Core","doc":"https://help.sap.com/docs/sap-ai-core","models":{"sonar-pro":{"id":"sonar-pro","name":"sonar-pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"gemini-3.1-flash-lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"gpt-5.6-terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"openai/gpt-5.6-terra"},"anthropic--claude-3-haiku":{"id":"anthropic--claude-3-haiku","name":"anthropic--claude-3-haiku","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-08-31","release_date":"2024-03-13","last_updated":"2024-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":4096},"cost":{"input":0.25,"output":1.25,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"anthropic/claude-3-haiku-20240307"},"anthropic--claude-4.6-sonnet":{"id":"anthropic--claude-4.6-sonnet","name":"anthropic--claude-4.6-sonnet","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"sonar":{"id":"sonar","name":"sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-6.1-sol"},"anthropic--claude-4.7-opus":{"id":"anthropic--claude-4.7-opus","name":"anthropic--claude-4.7-opus","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic--claude-4.6-opus":{"id":"anthropic--claude-4.6-opus","name":"anthropic--claude-4.6-opus","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"gemini-2.5-pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"google/gemini-2.5-pro"},"anthropic--claude-4-sonnet":{"id":"anthropic--claude-4-sonnet","name":"anthropic--claude-4-sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"nvidia--llama-3.2-nv-embedqa-1b":{"id":"nvidia--llama-3.2-nv-embedqa-1b","name":"nvidia--llama-3.2-nv-embedqa-1b","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":4096},"status":"deprecated"},"anthropic--claude-4.5-opus":{"id":"anthropic--claude-4.5-opus","name":"anthropic--claude-4.5-opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"cohere--command-a-reasoning":{"id":"cohere--command-a-reasoning","name":"cohere--command-a-reasoning","description":"Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]},{"type":"budget_tokens","min":1}],"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-reasoning-08-2025"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-6-luna"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"gpt-4.1-nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.4},"canonical_model_id":"openai/gpt-4.1-nano"},"anthropic--claude-4.5-haiku":{"id":"anthropic--claude-4.5-haiku","name":"anthropic--claude-4.5-haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"gpt-5.4":{"id":"gpt-5.4","name":"gpt-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"gpt-5.6-luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04}},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3.6-flash"},"amazon--nova-lite":{"id":"amazon--nova-lite","name":"amazon--nova-lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"amazon/nova-2-lite"},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Small, efficient OpenAI text embedding model","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":1536},"cost":{"input":0.02,"output":0},"canonical_model_id":"openai/text-embedding-3-small"},"nvidia--llama-nemotron-embed-1b":{"id":"nvidia--llama-nemotron-embed-1b","type":"embedding","name":"Llama Nemotron Embed 1B v2","description":"Multilingual text embedding model for retrieval and question answering pipelines","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"canonical_model_id":"nvidia/llama-nemotron-embed-1b-v2"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3.5-flash-lite"},"mistralai--mistral-small":{"id":"mistralai--mistral-small","name":"mistralai--mistral-small","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-2603"},"amazon--nova-micro":{"id":"amazon--nova-micro","name":"amazon--nova-micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.035,"output":0.14},"canonical_model_id":"amazon/nova-micro"},"sap-abap-1":{"id":"sap-abap-1","name":"sap-abap-1","description":"SAP-hosted model for ABAP code generation and enterprise development tasks","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-11-26","last_updated":"2025-11-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"status":"deprecated"},"gpt-5.2":{"id":"gpt-5.2","name":"gpt-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"gpt-4.1-mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"mistralai--mistral-medium":{"id":"mistralai--mistral-medium","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"canonical_model_id":"mistral/mistral-medium-2604"},"gpt-5-mini":{"id":"gpt-5-mini","name":"gpt-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-6-astra"},"amazon--nova-pro":{"id":"amazon--nova-pro","name":"amazon--nova-pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":8192},"cost":{"input":0.8,"output":3.2},"canonical_model_id":"amazon/nova-pro"},"mistralai--mistral-medium-instruct":{"id":"mistralai--mistral-medium-instruct","name":"mistralai--mistral-medium-instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-2505"},"anthropic--claude-4.5-sonnet":{"id":"anthropic--claude-4.5-sonnet","name":"anthropic--claude-4.5-sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"amazon--titan-embed-text":{"id":"amazon--titan-embed-text","name":"amazon--titan-embed-text","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-04-30","last_updated":"2024-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.02,"output":0}},"anthropic--claude-4-opus":{"id":"anthropic--claude-4-opus","name":"anthropic--claude-4-opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"sonar-deep-research":{"id":"sonar-deep-research","name":"sonar-deep-research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"gemini-2.5-flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Most capable OpenAI embedding model for English and non-English text","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":3072},"cost":{"input":0.13,"output":0},"canonical_model_id":"openai/text-embedding-3-large"},"gpt-5":{"id":"gpt-5","name":"gpt-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"gpt-5-nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3.8-flash"},"gpt-5.5":{"id":"gpt-5.5","name":"gpt-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"anthropic--claude-4.8-opus":{"id":"anthropic--claude-4.8-opus","name":"anthropic--claude-4.8-opus","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"gpt-5.6-sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"tiers":[{"input":8,"output":30,"cache_read":0.8,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8}},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-embedding":{"id":"gemini-embedding","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"canonical_model_id":"google/gemini-embedding-001"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"gemini-3.5-flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-4.1":{"id":"gpt-4.1","name":"gpt-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gemini-embedding-2":{"id":"gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"canonical_model_id":"google/gemini-embedding-2"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"gemini-2.5-flash-lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"}}},"submodel":{"id":"submodel","env":["SUBMODEL_INSTAGEN_ACCESS_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://llm.submodel.ai/v1","name":"submodel","doc":"https://submodel.gitbook.io","models":{"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.2,"output":0.3},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3 235B A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.2,"output":0.6},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8","name":"Qwen3 Coder 480B A35B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":75000,"output":163840},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":75000,"output":163840},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":75000,"output":163840},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"deepseek/deepseek-v3-0324"},"zai-org/GLM-4.5-FP8":{"id":"zai-org/GLM-4.5-FP8","name":"GLM 4.5 FP8","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"zhipuai/glm-4.5"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"GLM 4.5 Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"zhipuai/glm-4.5-air"}}},"edenai":{"id":"edenai","env":["EDENAI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.edenai.run/v3","name":"Eden AI","doc":"https://docs.edenai.co","models":{"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-latest":{"id":"xai/grok-latest","name":"Grok Latest (Grok 4.7)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4-fast-non-reasoning":{"id":"xai/grok-4-fast-non-reasoning","name":"X-Ai/Grok-4-Fast-Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05,"tiers":[{"input":0.4,"output":1,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-chat"},"azure/gpt-5.1-codex-mini":{"id":"azure/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"azure/gpt-5.2-codex":{"id":"azure/gpt-5.2-codex","name":"GPT-5.2 Codex (Azure)","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"azure/gpt-5.1-codex-max":{"id":"azure/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"azure/gpt-5.1-codex":{"id":"azure/gpt-5.1-codex","name":"GPT-5.1 Codex (Azure)","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"ovhcloud/gpt-oss-20b":{"id":"ovhcloud/gpt-oss-20b","name":"GPT OSS 20B (OVHcloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"ovhcloud/gpt-oss-120b":{"id":"ovhcloud/gpt-oss-120b","name":"GPT OSS 120B (OVHcloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.09,"output":0.47},"canonical_model_id":"openai/gpt-oss-120b"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B (Cerebras)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.35,"output":0.75,"cache_read":0.35},"canonical_model_id":"openai/gpt-oss-120b"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-5v-turbo":{"id":"zai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.6v":{"id":"zai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"groq/openai/gpt-oss-20b":{"id":"groq/openai/gpt-oss-20b","name":"GPT OSS 20B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"groq/openai/gpt-oss-safeguard-20b":{"id":"groq/openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B (Groq)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"groq/openai/gpt-oss-120b":{"id":"groq/openai/gpt-oss-120b","name":"GPT OSS 120B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"groq/meta-llama/llama-prompt-guard-2-86m":{"id":"groq/meta-llama/llama-prompt-guard-2-86m","name":"Llama Prompt Guard 2 86M (Groq)","description":"86M-parameter (mDeBERTa-base) multilingual classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"cost":{"input":0.04,"output":0.04,"cache_read":0.02},"canonical_model_id":"meta/llama-prompt-guard-2-86m"},"groq/meta-llama/llama-prompt-guard-2-22m":{"id":"groq/meta-llama/llama-prompt-guard-2-22m","name":"Llama Prompt Guard 2 22M (Groq)","description":"22M-parameter (DeBERTa-xsmall) English classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"cost":{"input":0.03,"output":0.03,"cache_read":0.015},"canonical_model_id":"meta/llama-prompt-guard-2-22m"},"nebius/Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"nebius/Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3 30B-A3B Instruct 2507 (Nebius)","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"nebius/Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"nebius/Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Nebius)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.2,"output":0.6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"nebius/google/gemma-3-27b-it":{"id":"nebius/google/gemma-3-27b-it","name":"Gemma 3 27B IT (Nebius)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":110000,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"google/gemma-3-27b-it"},"nebius/openai/gpt-oss-120b":{"id":"nebius/openai/gpt-oss-120b","name":"GPT OSS 120B (Nebius)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.15},"canonical_model_id":"openai/gpt-oss-120b"},"nebius/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"nebius/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Nebius)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"nebius/deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"nebius/deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813 (Nebius)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":979000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":1.32},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"nebius/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"nebius/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Nebius)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"nebius/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"nebius/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Nebius)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"nebius/meta-models/muse-spark-open":{"id":"nebius/meta-models/muse-spark-open","name":"Muse Spark Open (Nebius)","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1.25},"canonical_model_id":"meta/muse-spark-open"},"nebius/nvidia/Nemotron-3-Ultra-550b-a55b":{"id":"nebius/nvidia/Nemotron-3-Ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B (Nebius)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":1,"output":3,"cache_read":1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nebius/nvidia/nemotron-3-super-120b-a12b":{"id":"nebius/nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B (Nebius)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":0.9,"cache_read":0.3},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"vertex/gemini-3.8-flash@us":{"id":"vertex/gemini-3.8-flash@us","name":"Gemini 3.8 Flash (Vertex AI, US)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-3.7-flash@eu":{"id":"vertex/gemini-3.7-flash@eu","name":"Gemini 3.7 Flash (Vertex AI, EU)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"vertex/gemini-3.1-flash-lite":{"id":"vertex/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite (Vertex AI)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"vertex/gemini-flash-latest":{"id":"vertex/gemini-flash-latest","name":"Gemini Flash Latest (Gemini 3.8 Flash, Vertex AI)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-4-argon@us":{"id":"vertex/gemini-4-argon@us","name":"Gemini 4 Argon (Vertex AI, US)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-3.8-flash@eu":{"id":"vertex/gemini-3.8-flash@eu","name":"Gemini 3.8 Flash (Vertex AI, EU)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-3.1-flash-lite@us":{"id":"vertex/gemini-3.1-flash-lite@us","name":"Gemini 3.1 Flash Lite (Vertex AI, US)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"vertex/gemini-pro-latest":{"id":"vertex/gemini-pro-latest","name":"Gemini Pro Latest (Gemini 4 Argon, Vertex AI)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-3.6-flash":{"id":"vertex/gemini-3.6-flash","name":"Gemini 3.6 Flash (Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"vertex/gemini-3.6-flash@us":{"id":"vertex/gemini-3.6-flash@us","name":"Gemini 3.6 Flash (Vertex AI, US)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"vertex/gemini-3.6-flash@eu":{"id":"vertex/gemini-3.6-flash@eu","name":"Gemini 3.6 Flash (Vertex AI, EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"vertex/gemini-3.5-flash-lite":{"id":"vertex/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite (Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"vertex/gemini-3.7-flash@us":{"id":"vertex/gemini-3.7-flash@us","name":"Gemini 3.7 Flash (Vertex AI, US)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"vertex/gemini-3.1-flash-lite@eu":{"id":"vertex/gemini-3.1-flash-lite@eu","name":"Gemini 3.1 Flash Lite (Vertex AI, EU)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"vertex/gemini-4-argon":{"id":"vertex/gemini-4-argon","name":"Gemini 4 Argon (Vertex AI)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-4-argon@eu":{"id":"vertex/gemini-4-argon@eu","name":"Gemini 4 Argon (Vertex AI, EU)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-nano-banana-2.1":{"id":"vertex/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1 (Vertex AI)","description":"Image model for prompt-driven generation, multi-turn editing, and visual design workflows","family":"gemini-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"vertex/gemini-3.5-flash@us":{"id":"vertex/gemini-3.5-flash@us","name":"Gemini 3.5 Flash (Vertex AI, US)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"vertex/gemini-3.1-pro-preview":{"id":"vertex/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview (Vertex AI)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":0.25}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"vertex/gemini-3.7-flash":{"id":"vertex/gemini-3.7-flash","name":"Gemini 3.7 Flash (Vertex AI)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"vertex/gemini-3-flash-preview":{"id":"vertex/gemini-3-flash-preview","name":"Gemini 3 Flash Preview (Vertex AI)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"reasoning":3,"cache_read":0.05,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"vertex/gemini-3.5-flash-lite@us":{"id":"vertex/gemini-3.5-flash-lite@us","name":"Gemini 3.5 Flash Lite (Vertex AI, US)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"vertex/gemini-3-pro-image":{"id":"vertex/gemini-3-pro-image","type":"image","name":"Nano Banana Pro (Vertex AI)","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image"},"vertex/gemini-3.1-flash-image":{"id":"vertex/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2 (Vertex AI)","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"vertex/gemini-3.8-flash":{"id":"vertex/gemini-3.8-flash","name":"Gemini 3.8 Flash (Vertex AI)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-3.5-flash-lite@eu":{"id":"vertex/gemini-3.5-flash-lite@eu","name":"Gemini 3.5 Flash Lite (Vertex AI, EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"vertex/gemini-3.1-flash-lite-image":{"id":"vertex/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite (Vertex AI)","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"vertex/gemini-3.5-flash":{"id":"vertex/gemini-3.5-flash","name":"Gemini 3.5 Flash (Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"vertex/gemini-3.5-flash@eu":{"id":"vertex/gemini-3.5-flash@eu","name":"Gemini 3.5 Flash (Vertex AI, EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"vertex/gemini-2.5-flash-image":{"id":"vertex/gemini-2.5-flash-image","type":"image","name":"Nano Banana (Vertex AI)","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash-image"},"minimax/MiniMax-M2.7":{"id":"minimax/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/MiniMax-M2":{"id":"minimax/MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/MiniMax-M3":{"id":"minimax/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/MiniMax-M2.5":{"id":"minimax/MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/MiniMax-M2.1":{"id":"minimax/MiniMax-M2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/MiniMax-M3.1-Flash":{"id":"minimax/MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/c4ai-aya-expanse-32b":{"id":"cohere/c4ai-aya-expanse-32b","name":"Aya Expanse 32B","description":"Open multilingual model optimized for generation across 23 languages","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-10-24","last_updated":"2024-10-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"cohere/c4ai-aya-expanse-32b"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":132000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"cohere/command-a-03-2025":{"id":"cohere/command-a-03-2025","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":288000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-03-2025"},"arkanecloud/ministral-14b":{"id":"arkanecloud/ministral-14b","name":"Ministral 14B (Arkanecloud)","description":"Compact multimodal Mistral model for local assistants, edge agents, and efficient tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":262144},"cost":{"input":0.15,"output":0.15,"cache_read":0.15},"canonical_model_id":"mistral/ministral-14b"},"arkanecloud/Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"arkanecloud/Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Arkanecloud)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":16384},"cost":{"input":0.25,"output":1,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"arkanecloud/openai/gpt-oss-20b":{"id":"arkanecloud/openai/gpt-oss-20b","name":"GPT OSS 20B (Arkanecloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.25,"cache_read":0.05},"canonical_model_id":"openai/gpt-oss-20b"},"arkanecloud/openai/gpt-oss-120b":{"id":"arkanecloud/openai/gpt-oss-120b","name":"GPT OSS 120B (Arkanecloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-oss-120b"},"arkanecloud/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"arkanecloud/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Arkanecloud)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.075},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"arkanecloud/meta-llama/Llama-3.1-8B-Instruct":{"id":"arkanecloud/meta-llama/Llama-3.1-8B-Instruct","name":"Llama-3.1-8B-Instruct (Arkanecloud)","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.15,"output":0.15,"cache_read":0.03},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"arkanecloud/meta-llama/Llama-3.3-70B-Instruct":{"id":"arkanecloud/meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct (Arkanecloud)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.7,"output":0.7,"cache_read":0.2},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"arkanecloud/meta-llama/Llama-3.1-70B-Instruct":{"id":"arkanecloud/meta-llama/Llama-3.1-70B-Instruct","name":"Llama-3.1-70B-Instruct (Arkanecloud)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.8,"output":0.8,"cache_read":0.16},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"arkanecloud/nvidia/nemotron-3-ultra-550b-a55b":{"id":"arkanecloud/nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B (Arkanecloud)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512288,"output":128000},"cost":{"input":0.6,"output":3.5,"cache_read":0.2},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"arkanecloud/nvidia/nemotron-3-super-120b-a12b":{"id":"arkanecloud/nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B (Arkanecloud)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"cost":{"input":0.4,"output":0.8,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest (Gemini 3.8 Flash)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-pro-latest":{"id":"google/gemini-pro-latest","name":"Gemini Pro Latest (Gemini 4 Argon)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-robotics-er-2-preview":{"id":"google/gemini-robotics-er-2-preview","name":"Gemini Robotics-ER 2 Preview","description":"Embodied reasoning model for robotics: spatial understanding, video progress, and multi-robot orchestration","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","pdf","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":1,"output":5,"reasoning":5,"cache_read":0.1,"input_audio":1},"canonical_model_id":"google/gemini-robotics-er-2-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Image model for prompt-driven generation, multi-turn editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"reasoning":3,"cache_read":0.05,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image"},"google/deep-research-max-preview-04-2026":{"id":"google/deep-research-max-preview-04-2026","name":"Deep Research Max Preview","description":"Maximum-comprehensiveness agentic researcher for multi-step investigation, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/deep-research-max-preview-04-2026"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/deep-research-preview-04-2026":{"id":"google/deep-research-preview-04-2026","name":"Gemini Deep Research Preview","description":"Agentic model for autonomous multi-step research, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/deep-research-preview-04-2026"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["audio","image","text","video"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash-image"},"greenpt/deepseek-v4.1-flash":{"id":"greenpt/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Greenpt)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.246246,"output":1.231229,"cache_read":0.012312},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"greenpt/llama-3.3-70b-instruct":{"id":"greenpt/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct (Greenpt)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":1.231229,"output":1.231229},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"greenpt/qwen3-235b-a22b-instruct-2507":{"id":"greenpt/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Greenpt)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":250000,"output":16384},"cost":{"input":1.007369,"output":3.022108},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"greenpt/deepseek-v4-flash-0731":{"id":"greenpt/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Greenpt)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.156702,"output":0.391755,"cache_read":0.044772},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"greenpt/gpt-oss-120b":{"id":"greenpt/gpt-oss-120b","name":"GPT OSS 120B (Greenpt)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.22386,"output":0.78351},"canonical_model_id":"openai/gpt-oss-120b"},"mistral/codestral-2508":{"id":"mistral/codestral-2508","name":"Codestral 25.08","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"mistral/mistral-medium-latest":{"id":"mistral/mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-latest"},"mistral/mistral-large-2407":{"id":"mistral/mistral-large-2407","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-2407"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/voxtral-small-latest":{"id":"mistral/voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"mistral/voxtral-small-latest"},"mistral/ministral-14b-2512":{"id":"mistral/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b-2512"},"mistral/devstral-2512":{"id":"mistral/devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistral/mistral-medium-3":{"id":"mistral/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3"},"mistral/mistral-medium-2604":{"id":"mistral/mistral-medium-2604","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-2604"},"mistral/ministral-3b-2512":{"id":"mistral/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3b-2512"},"mistral/codestral-latest":{"id":"mistral/codestral-latest","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-latest"},"mistral/mistral-small-latest":{"id":"mistral/mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-latest"},"mistral/magistral-medium-latest":{"id":"mistral/magistral-medium-latest","name":"Magistral Medium (latest)","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-03-17","last_updated":"2025-03-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":16384},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/magistral-medium-latest"},"mistral/mistral-medium-3-5":{"id":"mistral/mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistral/mistral-medium-2505":{"id":"mistral/mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-2505"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/mistral-small-2603":{"id":"mistral/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistral/devstral-medium-latest":{"id":"mistral/devstral-medium-latest","name":"Devstral 2 (latest)","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-medium-latest"},"mistral/ministral-8b-2512":{"id":"mistral/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-8b-2512"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":150,"output":600},"canonical_model_id":"openai/o1-pro"},"openai/gpt-5-search-api":{"id":"openai/gpt-5-search-api","type":"chat","name":"GPT-5 Search API","description":"GPT-5 model trained for web search in the Chat Completions API","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":100000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-search-api"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-mini-latest":{"id":"openai/gpt-mini-latest","name":"GPT Mini Latest (GPT-5.6 Luna)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":8192},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-latest":{"id":"openai/gpt-latest","name":"GPT Latest (GPT-6 Astra)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-pro-latest":{"id":"openai/gpt-pro-latest","name":"GPT Pro Latest (GPT-5.6 Sol)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","pdf","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"cloudflare/@cf/aisingapore/gemma-sea-lion-v4-27b-it":{"id":"cloudflare/@cf/aisingapore/gemma-sea-lion-v4-27b-it","name":"Gemma-SEA-LION-v4-27B-IT (Cloudflare)","description":"Gemma 3 27B tuned by AI Singapore for Southeast Asian languages and instruction following","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"aisingapore/gemma-sea-lion-v4-27b-it"},"cloudflare/@cf/openai/gpt-oss-20b":{"id":"cloudflare/@cf/openai/gpt-oss-20b","name":"GPT OSS 20B (Cloudflare)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.2,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"cloudflare/@cf/openai/gpt-oss-120b":{"id":"cloudflare/@cf/openai/gpt-oss-120b","name":"GPT OSS 120B (Cloudflare)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"cloudflare/@cf/deepseek-ai/deepseek-v4.1-flash-1017":{"id":"cloudflare/@cf/deepseek-ai/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Cloudflare)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"cloudflare/@cf/deepseek-ai/deepseek-v4-pro-0813":{"id":"cloudflare/@cf/deepseek-ai/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Cloudflare)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"cloudflare/@cf/deepseek-ai/deepseek-v4-flash-0731":{"id":"cloudflare/@cf/deepseek-ai/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Cloudflare)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"cloudflare/@cf/meta/llama-guard-3-8b":{"id":"cloudflare/@cf/meta/llama-guard-3-8b","name":"Llama-Guard-3-8B (Cloudflare)","description":"Llama 3.1-based safety classifier for moderating prompts and model responses","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.484,"output":0.03},"canonical_model_id":"meta/llama-guard-3-8b"},"cloudflare/@cf/meta/llama-3.2-3b-instruct":{"id":"cloudflare/@cf/meta/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct (Cloudflare)","description":"Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":131072},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"cloudflare/@cf/meta/llama-3.2-1b-instruct":{"id":"cloudflare/@cf/meta/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct (Cloudflare)","description":"Compact instruction-tuned Llama model for multilingual dialogue and on-device assistants","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":32000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"cloudflare/@cf/mistralai/mistral-small-3.1-24b-instruct":{"id":"cloudflare/@cf/mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B (Cloudflare)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":102400},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"cloudflare/@cf/qwen/qwen2.5-coder-32b-instruct":{"id":"cloudflare/@cf/qwen/qwen2.5-coder-32b-instruct","name":"Qwen2.5-Coder-32B-Instruct (Cloudflare)","description":"Open coding-focused Qwen model for code generation, repair, and repository reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"cloudflare/@cf/zai-org/glm-4.7-flash":{"id":"cloudflare/@cf/zai-org/glm-4.7-flash","name":"GLM-4.7-Flash (Cloudflare)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"amazon/qwen.qwen3-coder-next":{"id":"amazon/qwen.qwen3-coder-next","name":"Qwen3 Coder Next (Amazon Bedrock)","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"amazon/amazon.nova-lite-v1:0":{"id":"amazon/amazon.nova-lite-v1:0","name":"Nova Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015},"canonical_model_id":"amazon/nova-lite"},"amazon/amazon.nova-pro-v1:0":{"id":"amazon/amazon.nova-pro-v1:0","name":"Nova Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2},"canonical_model_id":"amazon/nova-pro"},"amazon/google.gemma-3-12b-it@us":{"id":"amazon/google.gemma-3-12b-it@us","name":"Gemma 3 12B IT (Amazon Bedrock, US)","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"amazon/amazon.nova-micro-v1:0@us":{"id":"amazon/amazon.nova-micro-v1:0@us","name":"Nova Micro (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875},"canonical_model_id":"amazon/nova-micro"},"amazon/google.gemma-3-12b-it":{"id":"amazon/google.gemma-3-12b-it","name":"Gemma 3 12B IT (Amazon Bedrock)","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"amazon/moonshotai.kimi-k2.5":{"id":"amazon/moonshotai.kimi-k2.5","name":"Kimi K2.5 (Amazon Bedrock)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"amazon/zai.glm-4.7-flash@us":{"id":"amazon/zai.glm-4.7-flash@us","name":"GLM-4.7-Flash (Amazon Bedrock, US)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"amazon/mistral.pixtral-large-2502-v1:0":{"id":"amazon/mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02) (Amazon Bedrock)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"amazon/moonshot.kimi-k2-thinking":{"id":"amazon/moonshot.kimi-k2-thinking","name":"Kimi K2 Thinking (Amazon Bedrock)","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"amazon/google.gemma-3-27b-it":{"id":"amazon/google.gemma-3-27b-it","name":"Gemma 3 27B IT (Amazon Bedrock)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.23,"output":0.38},"canonical_model_id":"google/gemma-3-27b-it"},"amazon/amazon.nova-micro-v1:0":{"id":"amazon/amazon.nova-micro-v1:0","name":"Nova Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875},"canonical_model_id":"amazon/nova-micro"},"amazon/mistral.pixtral-large-2502-v1:0@us":{"id":"amazon/mistral.pixtral-large-2502-v1:0@us","name":"Pixtral Large (25.02) (Amazon Bedrock, US)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"amazon/google.gemma-3-4b-it":{"id":"amazon/google.gemma-3-4b-it","name":"Gemma 3 4B IT (Amazon Bedrock)","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"amazon/openai.gpt-oss-safeguard-20b@us":{"id":"amazon/openai.gpt-oss-safeguard-20b@us","name":"GPT OSS Safeguard 20B (Amazon Bedrock, US)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"amazon/amazon.nova-pro-v1:0@us":{"id":"amazon/amazon.nova-pro-v1:0@us","name":"Nova Pro (US)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2},"canonical_model_id":"amazon/nova-pro"},"amazon/openai.gpt-oss-safeguard-20b":{"id":"amazon/openai.gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B (Amazon Bedrock)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"amazon/zai.glm-4.7-flash":{"id":"amazon/zai.glm-4.7-flash","name":"GLM-4.7-Flash (Amazon Bedrock)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"amazon/mistral.voxtral-small-24b-2507@us":{"id":"amazon/mistral.voxtral-small-24b-2507@us","name":"Voxtral Small 24B 2507 (Amazon Bedrock, US)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"amazon/amazon.nova-lite-v1:0@us":{"id":"amazon/amazon.nova-lite-v1:0@us","name":"Nova Lite (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015},"canonical_model_id":"amazon/nova-lite"},"amazon/mistral.voxtral-mini-3b-2507":{"id":"amazon/mistral.voxtral-mini-3b-2507","name":"Voxtral Mini 3B 2507 (Amazon Bedrock)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.04,"output":0.04},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"amazon/google.gemma-3-4b-it@us":{"id":"amazon/google.gemma-3-4b-it@us","name":"Gemma 3 4B IT (Amazon Bedrock, US)","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"amazon/mistral.voxtral-mini-3b-2507@us":{"id":"amazon/mistral.voxtral-mini-3b-2507@us","name":"Voxtral Mini 3B 2507 (Amazon Bedrock, US)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.04,"output":0.04},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"amazon/mistral.voxtral-small-24b-2507":{"id":"amazon/mistral.voxtral-small-24b-2507","name":"Voxtral Small 24B 2507 (Amazon Bedrock)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"amazon/google.gemma-3-27b-it@us":{"id":"amazon/google.gemma-3-27b-it@us","name":"Gemma 3 27B IT (Amazon Bedrock, US)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.23,"output":0.38},"canonical_model_id":"google/gemma-3-27b-it"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1 (Meta)","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder (Meta)","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2 (Meta)","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3 (Meta)","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"fireworks_ai/gpt-oss-120b":{"id":"fireworks_ai/gpt-oss-120b","name":"GPT OSS 120B (Fireworks AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.014},"canonical_model_id":"openai/gpt-oss-120b"},"fireworks_ai/accounts/fireworks/models/muse-spark-open":{"id":"fireworks_ai/accounts/fireworks/models/muse-spark-open","name":"Muse Spark Open (Fireworks AI)","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"fireworks_ai/accounts/fireworks/models/inkling":{"id":"fireworks_ai/accounts/fireworks/models/inkling","name":"Inkling (Fireworks AI)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"fireworks_ai/accounts/fireworks/models/gpt-oss-120b":{"id":"fireworks_ai/accounts/fireworks/models/gpt-oss-120b","name":"GPT OSS 120B (Fireworks AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"},"fireworks_ai/accounts/fireworks/models/ember-1":{"id":"fireworks_ai/accounts/fireworks/models/ember-1","name":"Ember-1 (Fireworks AI)","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3.1-pro":{"id":"moonshot/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"flexai/DeepSeek-V4-Flash-0731":{"id":"flexai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (FlexAI)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.119,"output":0.238,"cache_read":0.0178},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"flexai/gpt-oss-20b":{"id":"flexai/gpt-oss-20b","name":"GPT OSS 20B (FlexAI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.02,"output":0.1,"cache_read":0.003},"canonical_model_id":"openai/gpt-oss-20b"},"flexai/DeepSeek-V4.1-Flash":{"id":"flexai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (FlexAI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.103,"output":0.715,"cache_read":0.0154},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"flexai/DeepSeek-V4-Flash-0731@eu":{"id":"flexai/DeepSeek-V4-Flash-0731@eu","name":"DeepSeek V4 Flash 0731 (FlexAI, EU)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.119,"output":0.238,"cache_read":0.0178},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"flexai/Step-3.7-Flash":{"id":"flexai/Step-3.7-Flash","name":"Step 3.7 Flash (FlexAI)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.03},"canonical_model_id":"stepfun/step-3.7-flash"},"flexai/gpt-oss-120b":{"id":"flexai/gpt-oss-120b","name":"GPT OSS 120B (FlexAI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.0075},"canonical_model_id":"openai/gpt-oss-120b"},"flexai/Muse-Glimmer-30B":{"id":"flexai/Muse-Glimmer-30B","name":"Muse Glimmer 30B (FlexAI)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.045},"canonical_model_id":"meta/muse-glimmer-30b"},"flexai/DeepSeek-V4.1-Flash-1017":{"id":"flexai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (FlexAI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.0225},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/deepseek-v4.1-flash":{"id":"qwen/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Alibaba)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"qwen/qwen-mt-plus":{"id":"qwen/qwen-mt-plus","name":"Qwen MT Plus","description":"Flagship Qwen translation model with terminology control and support for 92 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-24","last_updated":"2025-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":8192},"cost":{"input":2.46,"output":7.37},"canonical_model_id":"alibaba/qwen-mt-plus"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/deepseek-v4.1-flash-1017":{"id":"qwen/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Alibaba)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"qwen/deepseek-v4-pro-0813":{"id":"qwen/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Alibaba)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.066},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"qwen/deepseek-v4-flash-0731":{"id":"qwen/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Alibaba)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen-max":{"id":"qwen/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4,"cache_read":0.32},"canonical_model_id":"alibaba/qwen-max"},"scaleway/gemma-3-27b-it":{"id":"scaleway/gemma-3-27b-it","name":"Gemma 3 27B IT (Scaleway)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":131072},"cost":{"input":0.287125,"output":0.57425},"canonical_model_id":"google/gemma-3-27b-it"},"scaleway/deepseek-v4.1-flash":{"id":"scaleway/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Scaleway)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":384000},"cost":{"input":0.67158,"output":2.014739,"cache_read":0.134316},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"scaleway/llama-3.3-70b-instruct":{"id":"scaleway/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct (Scaleway)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":1.007369,"output":1.007369},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"scaleway/qwen3-coder-30b-a3b-instruct":{"id":"scaleway/qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct (Scaleway)","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":65536},"cost":{"input":0.2245,"output":0.898},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"scaleway/qwen3-235b-a22b-instruct-2507":{"id":"scaleway/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Scaleway)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":250000,"output":16384},"cost":{"input":0.839475,"output":2.518424},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"scaleway/deepseek-v4-flash-0731":{"id":"scaleway/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Scaleway)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":384000},"cost":{"input":0.44772,"output":0.89544,"cache_read":0.089544},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"scaleway/gpt-oss-120b":{"id":"scaleway/gpt-oss-120b","name":"GPT OSS 120B (Scaleway)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.167895,"output":0.67158},"canonical_model_id":"openai/gpt-oss-120b"},"together_ai/reflection/Beam-1.0-501B-A23B":{"id":"together_ai/reflection/Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B (Together AI)","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":1.2,"cache_read":0.04},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"together_ai/thinkingmachines/Inkling":{"id":"together_ai/thinkingmachines/Inkling","name":"Inkling (Together AI)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"together_ai/openai/gpt-oss-120b":{"id":"together_ai/openai/gpt-oss-120b","name":"GPT OSS 120B (Together AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"together_ai/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"together_ai/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Together AI)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"together_ai/deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"together_ai/deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813 (Together AI)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"together_ai/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"together_ai/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Together AI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"together_ai/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"together_ai/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Together AI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"together_ai/meta-models/Muse-Glimmer-30B":{"id":"together_ai/meta-models/Muse-Glimmer-30B","name":"Muse Glimmer 30B (Together AI)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"tencent/deepseek-v4.1-flash":{"id":"tencent/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Tencent)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3 (Tencent)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"tencent/deepseek-v4-pro-0813":{"id":"tencent/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Tencent)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"tencent/step-5-preview":{"id":"tencent/step-5-preview","name":"Step 5 Preview (Tencent)","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview (Tencent)","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-fable-latest":{"id":"anthropic/claude-fable-latest","name":"Claude Fable Latest (Claude Fable 5.1)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-sonnet-latest":{"id":"anthropic/claude-sonnet-latest","name":"Claude Sonnet Latest (Claude Sonnet 5.5)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-haiku-latest":{"id":"anthropic/claude-haiku-latest","name":"Claude Haiku Latest (Claude Haiku 5.5)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-latest":{"id":"anthropic/claude-opus-latest","name":"Claude Opus Latest (Claude Opus 5.5)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"sference/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"sference/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Sference)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"sference/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"sference/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Sference)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"perplexityai/sonar-pro":{"id":"perplexityai/sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"perplexityai/sonar-reasoning-pro":{"id":"perplexityai/sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"perplexityai/sonar":{"id":"perplexityai/sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127072,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"perplexityai/sonar-deep-research":{"id":"perplexityai/sonar-deep-research","name":"Sonar Deep Research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"},"greenference/deepseek-v4.1-flash":{"id":"greenference/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Greenference)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.22,"output":0.88,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"greenference/gpt-oss-20b":{"id":"greenference/gpt-oss-20b","name":"GPT OSS 20B (Greenference)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.009,"output":0.045,"cache_read":0.0045},"canonical_model_id":"openai/gpt-oss-20b"},"databricks/databricks-gpt-oss-20b@eu":{"id":"databricks/databricks-gpt-oss-20b@eu","name":"GPT OSS 20B (Databricks, EU)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.30002,"cache_read":0.007,"cache_write":0.07},"canonical_model_id":"openai/gpt-oss-20b"},"databricks/databricks-gpt-oss-20b":{"id":"databricks/databricks-gpt-oss-20b","name":"GPT OSS 20B (Databricks)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.30002,"cache_read":0.007,"cache_write":0.07},"canonical_model_id":"openai/gpt-oss-20b"},"databricks/databricks-gpt-oss-120b@eu":{"id":"databricks/databricks-gpt-oss-120b@eu","name":"GPT OSS 120B (Databricks, EU)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15001,"output":0.59997,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"openai/gpt-oss-120b"},"databricks/databricks-qwen3-next-80b-a3b-instruct@eu":{"id":"databricks/databricks-qwen3-next-80b-a3b-instruct@eu","name":"Qwen3-Next 80B-A3B Instruct (Databricks, EU)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15001,"output":1.20001,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"databricks/databricks-gpt-oss-120b":{"id":"databricks/databricks-gpt-oss-120b","name":"GPT OSS 120B (Databricks)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15001,"output":0.59997,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"openai/gpt-oss-120b"},"databricks/databricks-deepseek-v4-pro-0813":{"id":"databricks/databricks-deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Databricks)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.959999,"cache_read":0.132,"cache_write":1.31999},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"databricks/databricks-deepseek-v4-flash-0731":{"id":"databricks/databricks-deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Databricks)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.014,"cache_write":0.14},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"databricks/databricks-inkling":{"id":"databricks/databricks-inkling","name":"Inkling (Databricks)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1048576},"cost":{"input":1,"output":4.05,"cache_read":0.1,"cache_write":1.00002},"canonical_model_id":"thinkingmachines/inkling"},"databricks/databricks-qwen3-next-80b-a3b-instruct":{"id":"databricks/databricks-qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct (Databricks)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15001,"output":1.20001,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"tensorx/deepseek/deepseek-v4.1-flash":{"id":"tensorx/deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (TensorX)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.125},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"tensorx/deepseek/deepseek-v4.1-flash-1017":{"id":"tensorx/deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (TensorX)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.125},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"tensorx/deepseek/deepseek-v4-pro-0813":{"id":"tensorx/deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (TensorX)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":2,"output":4,"cache_read":0.5},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"tensorx/deepseek/deepseek-v4-flash-0731":{"id":"tensorx/deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (TensorX)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.25,"output":0.3,"cache_read":0.0625},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"tensorx/deepseek/deepseek-r1-0528":{"id":"tensorx/deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528 (TensorX)","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":32768},"cost":{"input":0.66,"output":2.6,"cache_read":0.165},"canonical_model_id":"deepseek/deepseek-r1-0528"},"tensorx/moonshotai/kimi-k2.5":{"id":"tensorx/moonshotai/kimi-k2.5","name":"Kimi K2.5 (TensorX)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.8,"cache_read":0.125},"canonical_model_id":"moonshotai/kimi-k2.5"},"infomaniak/mistralai/Ministral-3-14B-Instruct-2512":{"id":"infomaniak/mistralai/Ministral-3-14B-Instruct-2512","name":"Ministral 3 14B (Infomaniak)","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":262144},"cost":{"input":0.33579,"output":0.44772},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"ionos/Qwen/Qwen3-Coder-Next":{"id":"ionos/Qwen/Qwen3-Coder-Next","name":"Qwen3 Coder Next (IONOS)","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":65536},"cost":{"input":0.167895,"output":0.89544},"canonical_model_id":"alibaba/qwen3-coder-next"},"ionos/openai/gpt-oss-120b":{"id":"ionos/openai/gpt-oss-120b","name":"GPT OSS 120B (IONOS)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.167895,"output":0.727545},"canonical_model_id":"openai/gpt-oss-120b"},"ionos/meta-llama/Llama-3.3-70B-Instruct":{"id":"ionos/meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct (IONOS)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.727545,"output":0.727545},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"deepinfra/nemotron-3-ultra-550b-a55b":{"id":"deepinfra/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B (Deep Infra)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.5,"output":2.2,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"deepinfra/nemotron-3-super-120b-a12b":{"id":"deepinfra/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B (Deep Infra)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.085,"output":0.4},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"deepinfra/Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"deepinfra/Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3 VL 30B A3B Instruct (Deep Infra)","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"deepinfra/Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"deepinfra/Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Deep Infra)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3 235B-A22B Thinking 2507 (Deep Infra)","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.23,"output":2.3,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"deepinfra/Qwen/Qwen3-VL-235B-A22B-Instruct":{"id":"deepinfra/Qwen/Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct (Deep Infra)","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen3-Coder 480B-A35B Instruct (Deep Infra)","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"deepinfra/Qwen/Qwen2.5-7B-Instruct":{"id":"deepinfra/Qwen/Qwen2.5-7B-Instruct","name":"Qwen2.5 7B Instruct (Deep Infra)","description":"Small open Qwen2.5 instruction model for chat, structured output, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.04,"output":0.1},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"deepinfra/Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"deepinfra/Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next 80B-A3B Instruct (Deep Infra)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"deepinfra/Qwen/Qwen3-Max":{"id":"deepinfra/Qwen/Qwen3-Max","name":"Qwen3 Max (Deep Infra)","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1.2,"output":6,"cache_read":0.24},"canonical_model_id":"alibaba/qwen3-max"},"deepinfra/Gryphe/MythoMax-L2-13b":{"id":"deepinfra/Gryphe/MythoMax-L2-13b","name":"MythoMax L2 13B (Deep Infra)","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":3200},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"gryphe/mythomax-l2-13b"},"deepinfra/thinkingmachines/Inkling":{"id":"deepinfra/thinkingmachines/Inkling","name":"Inkling (Deep Infra)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.95,"output":4.05,"cache_read":0.16},"canonical_model_id":"thinkingmachines/inkling"},"deepinfra/thinkingmachines/Inkling-Small":{"id":"deepinfra/thinkingmachines/Inkling-Small","name":"Inkling Small (Deep Infra)","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"deepinfra/ibm-granite/granite-4.2-8b":{"id":"deepinfra/ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B (Deep Infra)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"deepinfra/ByteDance/Seed-2.0-mini":{"id":"deepinfra/ByteDance/Seed-2.0-mini","name":"Seed 2.0 Mini (Deep Infra)","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"deepinfra/ByteDance/Seed-2.0-code":{"id":"deepinfra/ByteDance/Seed-2.0-code","name":"Seed 2.0 Code (Deep Infra)","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"deepinfra/microsoft/WizardLM-2-8x22B":{"id":"deepinfra/microsoft/WizardLM-2-8x22B","name":"WizardLM 2 8x22B (Deep Infra)","description":"Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-04-15","last_updated":"2024-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8000},"cost":{"input":0.48,"output":0.48},"canonical_model_id":"microsoft/wizardlm-2-8x22b"},"deepinfra/google/gemma-3-27b-it":{"id":"deepinfra/google/gemma-3-27b-it","name":"Gemma 3 27B IT (Deep Infra)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"deepinfra/google/gemma-3-12b-it":{"id":"deepinfra/google/gemma-3-12b-it","name":"Gemma 3 12B IT (Deep Infra)","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"deepinfra/google/gemma-3-4b-it":{"id":"deepinfra/google/gemma-3-4b-it","name":"Gemma 3 4B IT (Deep Infra)","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-4b-it"},"deepinfra/openai/gpt-oss-20b":{"id":"deepinfra/openai/gpt-oss-20b","name":"GPT OSS 20B (Deep Infra)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.03,"output":0.14},"canonical_model_id":"openai/gpt-oss-20b"},"deepinfra/openai/gpt-oss-120b":{"id":"deepinfra/openai/gpt-oss-120b","name":"GPT OSS 120B (Deep Infra)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.037,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"deepinfra/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepinfra/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Deep Infra)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.06,"output":0.18,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepinfra/deepseek-ai/DeepSeek-V3":{"id":"deepinfra/deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3 (Deep Infra)","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":8192},"cost":{"input":0.32,"output":0.89},"canonical_model_id":"deepseek/deepseek-v3"},"deepinfra/deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepinfra/deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813 (Deep Infra)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepinfra/deepseek-ai/DeepSeek-R1-0528":{"id":"deepinfra/deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528 (Deep Infra)","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Deep Infra)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepinfra/deepseek-ai/DeepSeek-R1":{"id":"deepinfra/deepseek-ai/DeepSeek-R1","name":"DeepSeek-R1 (Deep Infra)","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.7,"output":2.4},"canonical_model_id":"deepseek/deepseek-r1"},"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Deep Infra)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepinfra/deepseek-ai/DeepSeek-V3-0324":{"id":"deepinfra/deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324 (Deep Infra)","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.24,"output":0.9,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepinfra/stepfun-ai/Step-3.5-Flash":{"id":"deepinfra/stepfun-ai/Step-3.5-Flash","name":"Step 3.5 Flash (Deep Infra)","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.09,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"deepinfra/stepfun-ai/Step-3.7-Flash":{"id":"deepinfra/stepfun-ai/Step-3.7-Flash","name":"Step 3.7 Flash (Deep Infra)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"deepinfra/stepfun-ai/Step-5-Preview":{"id":"deepinfra/stepfun-ai/Step-5-Preview","name":"Step 5 Preview (Deep Infra)","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"deepinfra/meta-models/Muse-Glimmer-30B":{"id":"deepinfra/meta-models/Muse-Glimmer-30B","name":"Muse Glimmer 30B (Deep Infra)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"deepinfra/moonshotai/Kimi-K2.5":{"id":"deepinfra/moonshotai/Kimi-K2.5","name":"Kimi K2.5 (Deep Infra)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"deepinfra/moonshotai/Kimi-K2-Instruct":{"id":"deepinfra/moonshotai/Kimi-K2-Instruct","name":"Kimi K2 Instruct (Deep Infra)","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"deepinfra/mistralai/Mixtral-8x7B-Instruct-v0.1":{"id":"deepinfra/mistralai/Mixtral-8x7B-Instruct-v0.1","name":"Mixtral 8x7B Instruct v0.1 (Deep Infra)","description":"Open sparse mixture-of-experts Mistral instruction model","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"mistral/mixtral-8x7b-instruct-v0.1"},"deepinfra/mistralai/Mistral-Small-24B-Instruct-2501":{"id":"deepinfra/mistralai/Mistral-Small-24B-Instruct-2501","name":"Mistral Small 3 (Deep Infra)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"deepinfra/NousResearch/Hermes-3-Llama-3.1-70B":{"id":"deepinfra/NousResearch/Hermes-3-Llama-3.1-70B","name":"Hermes 3 70B Instruct (Deep Infra)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.7,"output":0.7},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"deepinfra/NousResearch/Hermes-3-Llama-3.1-405B":{"id":"deepinfra/NousResearch/Hermes-3-Llama-3.1-405B","name":"Hermes 3 405B Instruct (Deep Infra)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1,"output":1},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"deepinfra/tencent/Hy4-preview":{"id":"deepinfra/tencent/Hy4-preview","name":"Hy4 preview (Deep Infra)","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"deepinfra/tencent/Hy3":{"id":"deepinfra/tencent/Hy3","name":"Hy3 (Deep Infra)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.13,"output":0.53,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"deepinfra/meta-llama/Llama-Guard-4-12B":{"id":"deepinfra/meta-llama/Llama-Guard-4-12B","name":"Llama Guard 4 12B (Deep Infra)","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"deepinfra/meta-llama/Llama-Guard-3-8B":{"id":"deepinfra/meta-llama/Llama-Guard-3-8B","name":"Llama-Guard-3-8B (Deep Infra)","description":"Llama 3.1-based safety classifier for moderating prompts and model responses","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.055,"output":0.055},"canonical_model_id":"meta/llama-guard-3-8b"},"deepinfra/meta-llama/Llama-3.3-70B-Instruct":{"id":"deepinfra/meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct (Deep Infra)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"deepinfra/meta-llama/Llama-3.2-3B-Instruct":{"id":"deepinfra/meta-llama/Llama-3.2-3B-Instruct","name":"Llama 3.2 3B Instruct (Deep Infra)","description":"Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.02,"output":0.02},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"deepinfra/meta-llama/Llama-3.2-11B-Vision-Instruct":{"id":"deepinfra/meta-llama/Llama-3.2-11B-Vision-Instruct","name":"Llama-3.2-11B-Vision-Instruct (Deep Infra)","description":"Open multimodal Llama model for image understanding, captioning, and visual QA","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.345,"output":0.345},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"deepinfra/nvidia/Nemotron-3-Nano-30B-A3B":{"id":"deepinfra/nvidia/Nemotron-3-Nano-30B-A3B","name":"Nemotron 3 Nano 30B A3B (Deep Infra)","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"deepinfra/nvidia/Llama-3.1-Nemotron-70B-Instruct":{"id":"deepinfra/nvidia/Llama-3.1-Nemotron-70B-Instruct","name":"Llama 3.1 Nemotron 70B Instruct (Deep Infra)","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.6,"output":0.6},"canonical_model_id":"nvidia/llama-3.1-nemotron-70b-instruct"},"deepinfra/zai-org/GLM-4.7-Flash":{"id":"deepinfra/zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash (Deep Infra)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"}}},"upstage":{"id":"upstage","env":["UPSTAGE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.upstage.ai/v1/solar","name":"Upstage","doc":"https://developers.upstage.ai/docs/apis/chat","models":{"solar-mini":{"id":"solar-mini","name":"solar-mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"solar-mini","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-06-12","last_updated":"2025-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"cost":{"input":0.15,"output":0.15}},"solar-pro3":{"id":"solar-pro3","name":"solar-pro3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.25,"output":0.25},"canonical_model_id":"upstage/solar-pro3"},"solar-pro4":{"id":"solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"upstage/solar-pro4"},"solar-pro2":{"id":"solar-pro2","name":"solar-pro2","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"cost":{"input":0.25,"output":0.25},"canonical_model_id":"upstage/solar-pro2"},"solar-mini4":{"id":"solar-mini4","name":"Solar Mini 4","description":"Upstage's compact, cost-efficient model for agentic use","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"upstage/solar-mini4"}}},"moonshotai-cn":{"id":"moonshotai-cn","env":["MOONSHOT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.moonshot.cn/v1","name":"Moonshot AI (China)","doc":"https://platform.moonshot.cn/docs/api/chat","models":{"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":2.9879,"output":14.9395,"cache_read":0.2988,"cache_write":2.9879},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.9711,"output":4.0337,"cache_read":0.1942},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.9437,"output":10.1671,"cache_read":0.1944,"cache_write":1.9437},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.9711,"output":4.0337,"cache_read":0.1643},"canonical_model_id":"moonshotai/kimi-k2.6"},"kimi-k3.1-pro-highspeed":{"id":"kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":3.8874,"output":20.3342,"cache_read":0.1944,"cache_write":3.8874},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code HighSpeed","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9421,"output":8.0673,"cache_read":0.3884},"canonical_model_id":"moonshotai/kimi-k2.7-code"}}},"google":{"id":"google","env":["GOOGLE_API_KEY","GOOGLE_GENERATIVE_AI_API_KEY","GEMINI_API_KEY"],"npm":"@ai-sdk/google","name":"Google","doc":"https://ai.google.dev/gemini-api/docs/models","models":{"gemini-omni-flash-preview":{"id":"gemini-omni-flash-preview","type":"video","name":"Gemini Omni Flash Preview","description":"Video generation and editing model for fast, conversational text- and image-to-video workflows","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":1.5,"output":17.5},"canonical_model_id":"google/gemini-omni-flash-preview"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5}},"gemini-flash-latest":{"id":"gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75}},"gemini-embedding-001":{"id":"gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"cost":{"input":0.15,"output":0}},"gemini-2.5-computer-use-preview-10-2025":{"id":"gemini-2.5-computer-use-preview-10-2025","name":"Gemini 2.5 Computer Use Preview 10-2025","description":"Specialized Gemini 2.5 model for browser-control agents that automate UI tasks","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-10-07","last_updated":"2025-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":1.25,"output":10,"tiers":[{"input":2.5,"output":15,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15}},"canonical_model_id":"google/gemini-2.5-computer-use-preview-10-2025"},"gemini-pro-latest":{"id":"gemini-pro-latest","name":"Gemini Pro Latest","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"gemini-3.1-pro-preview-customtools":{"id":"gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}}},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image-preview"},"lyria-3-clip-preview":{"id":"lyria-3-clip-preview","name":"Lyria 3 Clip Preview","description":"Music generation model for short 30-second clips, loops, and previews from text or image prompts","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-clip-preview"},"gemini-3.5-live-translate-preview":{"id":"gemini-3.5-live-translate-preview","type":"realtime","name":"Gemini 3.5 Live Translate Preview","description":"Low-latency audio-to-audio model for real-time speech translation across 70+ languages","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["audio"],"output":["audio","text"]},"open_weights":false,"limit":{"context":16384,"output":32768},"cost":{"input":3.5,"output":21,"input_audio":3.5,"output_audio":21},"canonical_model_id":"google/gemini-3.5-live-translate-preview"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768}},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"lyria-3-pro-preview":{"id":"lyria-3-pro-preview","name":"Lyria 3 Pro Preview","description":"Music generation model for full-length songs from text or images with vocals and structure","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-pro-preview"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Legacy model retained for compatibility with older integrations","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5}},"gemini-2.5-flash-preview-tts":{"id":"gemini-2.5-flash-preview-tts","type":"speech","name":"Gemini 2.5 Flash Preview TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-01","last_updated":"2025-05-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"cost":{"input":0.5,"output":10}},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}}},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1}},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image"},"deep-research-max-preview-04-2026":{"id":"deep-research-max-preview-04-2026","name":"Deep Research Max Preview (Apr-21-2026)","description":"Maximum-comprehensiveness agentic researcher for multi-step investigation, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/deep-research-max-preview-04-2026"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":60},"canonical_model_id":"google/gemini-3.1-flash-image"},"gemini-3.1-flash-tts-preview":{"id":"gemini-3.1-flash-tts-preview","type":"speech","name":"Gemini 3.1 Flash TTS Preview","description":"Low-latency speech generation with steerable prompts and expressive audio tags","family":"gemini-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-15","last_updated":"2026-04-15","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"cost":{"input":1,"output":20},"canonical_model_id":"google/gemini-3.1-flash-tts-preview"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768}},"deep-research-preview-04-2026":{"id":"deep-research-preview-04-2026","name":"Deep Research Preview (Apr-21-2026)","description":"Agentic model for autonomous multi-step research, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/deep-research-preview-04-2026"},"gemini-2.5-pro-preview-tts":{"id":"gemini-2.5-pro-preview-tts","type":"speech","name":"Gemini 2.5 Pro Preview TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-01","last_updated":"2025-05-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"cost":{"input":1,"output":20}},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"gemini-3.1-flash-lite-image":{"id":"gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":30},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"gemini-3.1-flash-image-preview":{"id":"gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":60}},"gemini-3.1-flash-live-preview":{"id":"gemini-3.1-flash-live-preview","type":"realtime","name":"Gemini 3.1 Flash Live Preview","description":"High-quality, low-latency Live API model for real-time dialogue and voice-first AI applications","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-26","last_updated":"2026-03-26","modalities":{"input":["text","image","video","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":0.75,"output":4.5,"input_audio":3,"output_audio":12},"canonical_model_id":"google/gemini-3.1-flash-live-preview"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5}},"gemini-flash-lite-latest":{"id":"gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30,"cache_read":0.075},"canonical_model_id":"google/gemini-2.5-flash-image"},"gemini-embedding-2":{"id":"gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1},"cost":{"input":0.2,"output":0,"input_audio":6.5},"canonical_model_id":"google/gemini-embedding-2"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3}}}},"siliconflow-cn":{"id":"siliconflow-cn","env":["SILICONFLOW_CN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.siliconflow.cn/v1","name":"SiliconFlow (China)","doc":"https://cloud.siliconflow.com/models","models":{"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen/Qwen3-30B-A3B-Instruct-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3.5-4B":{"id":"Qwen/Qwen3.5-4B","name":"Qwen/Qwen3.5-4B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-4b"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen/Qwen3-32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.59778991},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen/Qwen3-VL-30B-A3B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen/Qwen3.5-35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.05977899,"output":0.47823193,"tiers":[{"input":0.23911596,"output":1.91292771,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen/Qwen3.5-27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.08966849,"output":0.71734789,"tiers":[{"input":0.26900546,"output":2.15204368,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-VL-8B-Instruct":{"id":"Qwen/Qwen3-VL-8B-Instruct","name":"Qwen/Qwen3-VL-8B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.07472374,"output":0.29889496}},"Qwen/Qwen3-14B":{"id":"Qwen/Qwen3-14B","name":"Qwen/Qwen3-14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.07472374,"output":0.29889496}},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen/Qwen3-8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3-VL-30B-A3B-Thinking":{"id":"Qwen/Qwen3-VL-30B-A3B-Thinking","name":"Qwen/Qwen3-VL-30B-A3B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-11","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294}},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen/Qwen3.6-35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26900546,"output":1.61403276},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen2.5-7B-Instruct":{"id":"Qwen/Qwen2.5-7B-Instruct","name":"Qwen/Qwen2.5-7B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"Qwen/Qwen3-VL-32B-Thinking":{"id":"Qwen/Qwen3-VL-32B-Thinking","name":"Qwen/Qwen3-VL-32B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.14944748,"output":1.49447478}},"Qwen/Qwen3-VL-32B-Instruct":{"id":"Qwen/Qwen3-VL-32B-Instruct","name":"Qwen/Qwen3-VL-32B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.14944748,"output":0.59778991}},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen/Qwen3.5-9B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.07472374,"output":0.59778991,"tiers":[{"input":0.22417122,"output":1.79336973,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen/Qwen3-Coder-30B-A3B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"Qwen/Qwen2.5-72B-Instruct":{"id":"Qwen/Qwen2.5-72B-Instruct","name":"Qwen/Qwen2.5-72B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0.61721808,"output":0.61721808}},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen/Qwen3.5-122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.11955798,"output":0.95646386,"tiers":[{"input":0.29889496,"output":2.39115964,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"ByteDance-Seed/Seed-OSS-36B-Instruct":{"id":"ByteDance-Seed/Seed-OSS-36B-Instruct","name":"ByteDance-Seed/Seed-OSS-36B-Instruct","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-04","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.22417122,"output":0.59778991}},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"deepseek-ai/DeepSeek-V4-Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":393000},"cost":{"input":1.79336973,"output":3.58673947,"cache_read":0.14944748},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44834243,"output":1.3450273,"cache_read":0.04483424},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"deepseek-ai/DeepSeek-V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.29889496,"output":1.19557982,"cache_read":0.0298895},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-OCR":{"id":"deepseek-ai/DeepSeek-OCR","name":"deepseek-ai/DeepSeek-OCR","description":"OCR model for extracting structured text from documents and screenshots","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0}},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"deepseek-ai/DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":0.89668487,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"deepseek-ai/DeepSeek-R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":2.39115964,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"deepseek-ai/DeepSeek-V3.1-Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":1.79336973,"cache_read":0.05977899}},"stepfun-ai/Step-3.5-Flash":{"id":"stepfun-ai/Step-3.5-Flash","name":"stepfun-ai/Step-3.5-Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.3138397},"canonical_model_id":"stepfun/step-3.5-flash"},"inclusionAI/Ling-flash-2.0":{"id":"inclusionAI/Ling-flash-2.0","name":"inclusionAI/Ling-flash-2.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.59778991}},"PaddlePaddle/PaddleOCR-VL-1.5":{"id":"PaddlePaddle/PaddleOCR-VL-1.5","name":"PaddlePaddle/PaddleOCR-VL-1.5","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-29","last_updated":"2026-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0,"output":0}},"tencent/Hunyuan-A13B-Instruct":{"id":"tencent/Hunyuan-A13B-Instruct","name":"tencent/Hunyuan-A13B-Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.59778991}},"Pro/deepseek-ai/DeepSeek-V3":{"id":"Pro/deepseek-ai/DeepSeek-V3","name":"Pro/deepseek-ai/DeepSeek-V3","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.29889496,"output":1.19557982,"cache_read":0.0298895},"canonical_model_id":"deepseek/deepseek-v3-0324"},"Pro/deepseek-ai/DeepSeek-V3.2":{"id":"Pro/deepseek-ai/DeepSeek-V3.2","name":"Pro/deepseek-ai/DeepSeek-V3.2","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":0.89668487,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-v3.2"},"Pro/deepseek-ai/DeepSeek-R1":{"id":"Pro/deepseek-ai/DeepSeek-R1","name":"Pro/deepseek-ai/DeepSeek-R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":2.39115964,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-r1-0528"},"Pro/deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"Pro/deepseek-ai/DeepSeek-V3.1-Terminus","name":"Pro/deepseek-ai/DeepSeek-V3.1-Terminus","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":1.79336973,"cache_read":0.05977899}},"Pro/moonshotai/Kimi-K2.5":{"id":"Pro/moonshotai/Kimi-K2.5","name":"Pro/moonshotai/Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"status":"deprecated","cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"Pro/moonshotai/Kimi-K2.6":{"id":"Pro/moonshotai/Kimi-K2.6","name":"Pro/moonshotai/Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.97140861,"output":4.0350819,"cache_read":0.16439223},"canonical_model_id":"moonshotai/kimi-k2.6"},"Pro/zai-org/GLM-5.1":{"id":"Pro/zai-org/GLM-5.1","name":"Pro/zai-org/GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":205000},"cost":{"input":0.89668487,"output":3.58673947,"cache_read":0.19428172,"cache_write":0,"tiers":[{"input":1.19557982,"output":4.18452938,"cache_read":0.29889496,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"Pro/zai-org/GLM-5":{"id":"Pro/zai-org/GLM-5","name":"Pro/zai-org/GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":205000},"status":"deprecated","cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"zai-org/GLM-4.5-Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.89668487},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":1.19557982,"output":4.18452938,"cache_read":0.29889496,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"llmgateway-providers":{"id":"llmgateway-providers","env":["LLMGATEWAY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.llmgateway.io/v1","name":"LLM Gateway","doc":"https://llmgateway.io/docs","models":{"xai/grok-4-20-beta-0309-non-reasoning":{"id":"xai/grok-4-20-beta-0309-non-reasoning","name":"Grok 4.20 Beta Non-Reasoning (0309) (xAI)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4-7":{"id":"xai/grok-4-7","name":"Grok 4.7 (xAI)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4":{"id":"xai/grok-4","name":"Grok 4 (xAI)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":3,"output":15,"cache_read":0.75}},"xai/grok-4-3":{"id":"xai/grok-4-3","name":"Grok 4.3 (xAI)","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4-6":{"id":"xai/grok-4-6","name":"Grok 4.6 (xAI)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"xai/grok-build-0-1":{"id":"xai/grok-build-0-1","name":"Grok Build 0.1 (xAI)","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4-20-beta-0309-reasoning":{"id":"xai/grok-4-20-beta-0309-reasoning","name":"Grok 4.20 Beta Reasoning (0309) (xAI)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4-5":{"id":"xai/grok-4-5","name":"Grok 4.5 (xAI)","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"aws-mantle/gpt-5.6-terra":{"id":"aws-mantle/gpt-5.6-terra","name":"GPT-5.6 Terra (AWS Mantle)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":921600,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"openai/gpt-5.6-terra"},"aws-mantle/gpt-5.6-luna":{"id":"aws-mantle/gpt-5.6-luna","name":"GPT-5.6 Luna (AWS Mantle)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":921600,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275},"canonical_model_id":"openai/gpt-5.6-luna"},"aws-mantle/gpt-6-astra":{"id":"aws-mantle/gpt-6-astra","name":"GPT-6 Astra (AWS Mantle)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"aws-mantle/gpt-5.6-sol":{"id":"aws-mantle/gpt-5.6-sol","name":"GPT-5.6 Sol (AWS Mantle)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":921600,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5},"canonical_model_id":"openai/gpt-5.6-sol"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (DeepSeek)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro (DeepSeek)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":393216},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (DeepSeek)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"azure/gpt-5.1-codex-mini":{"id":"azure/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"azure/gpt-5.6-terra":{"id":"azure/gpt-5.6-terra","name":"GPT-5.6 Terra (Azure)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"azure/gpt-5.2-codex":{"id":"azure/gpt-5.2-codex","name":"GPT-5.2 Codex (Azure)","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"azure/gpt-6.1-sol":{"id":"azure/gpt-6.1-sol","name":"GPT-6.1 Sol (Azure)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"azure/gpt-6-luna":{"id":"azure/gpt-6-luna","name":"GPT-6 Luna (Azure)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"azure/gpt-5.4":{"id":"azure/gpt-5.4","name":"GPT-5.4 (Azure)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"azure/gpt-5.6-luna":{"id":"azure/gpt-5.6-luna","name":"GPT-5.6 Luna (Azure)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"azure/gpt-5.3-codex":{"id":"azure/gpt-5.3-codex","name":"GPT-5.3 Codex (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"azure/gpt-5.2":{"id":"azure/gpt-5.2","name":"GPT-5.2 (Azure)","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"azure/gpt-4.1-mini":{"id":"azure/gpt-4.1-mini","name":"GPT-4.1 Mini (Azure)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"azure/gpt-5-mini":{"id":"azure/gpt-5-mini","name":"GPT-5 Mini (Azure)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"azure/gpt-6-astra":{"id":"azure/gpt-6-astra","name":"GPT-6 Astra (Azure)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"azure/gpt-5.4-pro":{"id":"azure/gpt-5.4-pro","name":"GPT-5.4 Pro (Azure)","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"azure/gpt-5.1":{"id":"azure/gpt-5.1","name":"GPT-5.1 (Azure)","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"azure/gpt-5.4-nano":{"id":"azure/gpt-5.4-nano","name":"GPT-5.4 Nano (Azure)","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"azure/gpt-4o":{"id":"azure/gpt-4o","name":"GPT-4o (Azure)","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"azure/gpt-oss-120b":{"id":"azure/gpt-oss-120b","name":"GPT OSS 120B (Azure)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"azure/gpt-5":{"id":"azure/gpt-5","name":"GPT-5 (Azure)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"azure/gpt-5-nano":{"id":"azure/gpt-5-nano","name":"GPT-5 Nano (Azure)","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"azure/gpt-5.1-codex":{"id":"azure/gpt-5.1-codex","name":"GPT-5.1 Codex (Azure)","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-codex"},"azure/gpt-5.5":{"id":"azure/gpt-5.5","name":"GPT-5.5 (Azure)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"azure/gpt-5.6-sol":{"id":"azure/gpt-5.6-sol","name":"GPT-5.6 Sol (Azure)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"azure/gpt-6-sol":{"id":"azure/gpt-6-sol","name":"GPT-6 Sol (Azure)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"azure/gpt-4.1":{"id":"azure/gpt-4.1","name":"GPT-4.1 (Azure)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"azure/gpt-5.4-mini":{"id":"azure/gpt-5.4-mini","name":"GPT-5.4 Mini (Azure)","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max (Sakana AI)","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/fugu-ultra-v2.0":{"id":"sakana/fugu-ultra-v2.0","name":"Fugu Ultra v2.0 (Sakana AI)","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra (Sakana AI)","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"vertex-openai/grok-4-7":{"id":"vertex-openai/grok-4-7","name":"Grok 4.7 (Vertex AI (OpenAI-compatible))","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":524288},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"vertex-openai/grok-4-20-reasoning":{"id":"vertex-openai/grok-4-20-reasoning","name":"Grok 4.20 Reasoning (Vertex AI (OpenAI-compatible))","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"vertex-openai/grok-4-20-non-reasoning":{"id":"vertex-openai/grok-4-20-non-reasoning","name":"Grok 4.20 Non-Reasoning (Vertex AI (OpenAI-compatible))","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"vertex-openai/grok-4-6":{"id":"vertex-openai/grok-4-6","name":"Grok 4.6 (Vertex AI (OpenAI-compatible))","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"luminal/deepseek-v4.1-flash":{"id":"luminal/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (luminal)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.135,"output":0.54,"cache_read":0.0027},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"together-ai/deepseek-v4.1-flash":{"id":"together-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Together AI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"together-ai/glm-5.3":{"id":"together-ai/glm-5.3","name":"GLM-5.3 (Together AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"together-ai/kimi-k3":{"id":"together-ai/kimi-k3","name":"Kimi K3 (Together AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"together-ai/deepseek-v4-pro":{"id":"together-ai/deepseek-v4-pro","name":"DeepSeek V4 Pro (Together AI)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":163840},"cost":{"input":1.32,"output":3.96,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro"},"together-ai/glm-4.7":{"id":"together-ai/glm-4.7","name":"GLM-4.7 (Together AI)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0.45,"output":2},"canonical_model_id":"zhipuai/glm-4.7"},"together-ai/deepseek-v4.1-flash-1017":{"id":"together-ai/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Together AI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"together-ai/muse-glimmer-30b":{"id":"together-ai/muse-glimmer-30b","name":"Muse Glimmer 30B (Together AI)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"together-ai/minimax-m3":{"id":"together-ai/minimax-m3","name":"MiniMax M3 (Together AI)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"together-ai/qwen3.8-2.4t-a95b":{"id":"together-ai/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B (Together AI)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":909000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"together-ai/glm-5.3-flash":{"id":"together-ai/glm-5.3-flash","name":"GLM-5.3 Flash (Together AI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"together-ai/gpt-oss-120b":{"id":"together-ai/gpt-oss-120b","name":"GPT OSS 120B (Together AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"reka-ai/deepseek-v4.1-flash":{"id":"reka-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (reka-ai)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"reka-ai/glm-5.3":{"id":"reka-ai/glm-5.3","name":"GLM-5.3 (reka-ai)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.17,"output":3.96,"cache_read":0.234},"canonical_model_id":"zhipuai/glm-5.3"},"reka-ai/deepseek-v4-pro":{"id":"reka-ai/deepseek-v4-pro","name":"DeepSeek V4 Pro (reka-ai)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.2,"output":3.3,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro"},"reka-ai/reka-edge-2603":{"id":"reka-ai/reka-edge-2603","name":"Reka Edge 2603 (reka-ai)","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-09","last_updated":"2026-10-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":16384},"cost":{"input":0.1,"output":0.1}},"reka-ai/deepseek-v4.1-flash-1017":{"id":"reka-ai/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (reka-ai)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"reka-ai/glm-5.3-flash":{"id":"reka-ai/glm-5.3-flash","name":"GLM-5.3 Flash (reka-ai)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-5.3-flash"},"cerebras/llama-3.3-70b-instruct":{"id":"cerebras/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct (Cerebras)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.85,"output":1.2},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"cerebras/qwen3-235b-a22b-instruct-2507":{"id":"cerebras/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507 (Cerebras)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":8192},"cost":{"input":0.6,"output":1.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"cerebras/glm-4.7":{"id":"cerebras/glm-4.7","name":"GLM-4.7 (Cerebras)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":2.25,"output":2.75},"canonical_model_id":"zhipuai/glm-4.7"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B (Cerebras)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"cerebras/gemma-4-31b-it":{"id":"cerebras/gemma-4-31b-it","name":"Gemma 4 31B IT (Cerebras)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.99,"output":1.49},"canonical_model_id":"google/gemma-4-31b-it"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5 (Z AI)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131100},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3 (Z AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1 (Z AI)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"GLM-4.5V (Z AI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM-4.7 FlashX (Z AI)","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7 (Z AI)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5 (Z AI)","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-4-32b-0414-128k":{"id":"zai/glm-4-32b-0414-128k","name":"GLM-4 32B (0414-128k) (Z AI)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":0.1,"output":0.1}},"zai/glm-4.6v-flashx":{"id":"zai/glm-4.6v-flashx","name":"GLM-4.6V FlashX (Z AI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.04,"output":0.4,"cache_read":0.004}},"zai/glm-4.5-airx":{"id":"zai/glm-4.5-airx","name":"GLM-4.5 AirX (Z AI)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":1.1,"output":4.5,"cache_read":0.22}},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2 (Z AI)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3 Flash (Z AI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6 (Z AI)","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.6v":{"id":"zai/glm-4.6v","name":"GLM-4.6V (Z AI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5 Air (Z AI)","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai/glm-4.5-x":{"id":"zai/glm-4.5-x","name":"GLM-4.5 X (Z AI)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":2.2,"output":8.9,"cache_read":0.45}},"inference.net/glm-5.3":{"id":"inference.net/glm-5.3","name":"GLM-5.3 (Inference.net)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.9,"output":3,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-5.3"},"inference.net/llama-3.2-11b-instruct":{"id":"inference.net/llama-3.2-11b-instruct","name":"Llama 3.2 11B Instruct (Inference.net)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":0.07,"output":0.33}},"inference.net/deepseek-v4.1-flash-1017":{"id":"inference.net/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Inference.net)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.3,"output":1,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"inference.net/glm-5.3-flash":{"id":"inference.net/glm-5.3-flash","name":"GLM-5.3 Flash (Inference.net)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":0.09,"output":0.28,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-5.3-flash"},"alibaba/glm-5":{"id":"alibaba/glm-5","name":"GLM-5 (Alibaba Cloud)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"cost":{"input":0.573,"output":2.58},"canonical_model_id":"zhipuai/glm-5"},"alibaba/deepseek-v4.1-flash":{"id":"alibaba/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Alibaba Cloud)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"alibaba/glm-5.3":{"id":"alibaba/glm-5.3","name":"GLM-5.3 (Alibaba Cloud)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.28},"canonical_model_id":"zhipuai/glm-5.3"},"alibaba/kimi-k3":{"id":"alibaba/kimi-k3","name":"Kimi K3 (Alibaba Cloud)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"alibaba/qwen3.6-plus":{"id":"alibaba/qwen3.6-plus","name":"Qwen3.6 Plus (Alibaba Cloud)","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"alibaba/deepseek-v4-pro":{"id":"alibaba/deepseek-v4-pro","name":"DeepSeek V4 Pro (Alibaba Cloud)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":2.4,"output":4.8,"cache_read":0.2},"canonical_model_id":"deepseek/deepseek-v4-pro"},"alibaba/qwen3.7-flash":{"id":"alibaba/qwen3.7-flash","name":"Qwen3.7 Flash (Alibaba Cloud)","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.0375},"canonical_model_id":"alibaba/qwen3.7-flash"},"alibaba/qwen-plus-latest":{"id":"alibaba/qwen-plus-latest","name":"Qwen Plus Latest (Alibaba Cloud)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-09-09","last_updated":"2024-09-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.4,"output":1.2,"cache_read":0.08,"cache_write":0.5}},"alibaba/qwen3-coder-flash":{"id":"alibaba/qwen3-coder-flash","name":"Qwen3 Coder Flash (Alibaba Cloud)","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":997952,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"alibaba/qwen3.7-max":{"id":"alibaba/qwen3.7-max","name":"Qwen3.7 Max (Alibaba Cloud)","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"alibaba/kimi-k2.5":{"id":"alibaba/kimi-k2.5","name":"Kimi K2.5 (Alibaba Cloud)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.574,"output":3.011},"canonical_model_id":"moonshotai/kimi-k2.5"},"alibaba/qwen3.5-35b-a3b":{"id":"alibaba/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"alibaba/qwen3-vl-plus":{"id":"alibaba/qwen3-vl-plus","name":"Qwen3 VL Plus (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":1.6,"cache_read":0.04,"cache_write":0.25},"canonical_model_id":"alibaba/qwen3-vl-plus"},"alibaba/qwen-flash":{"id":"alibaba/qwen-flash","name":"Qwen Flash (Alibaba Cloud)","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.05,"output":0.4,"cache_read":0.01,"cache_write":0.0625},"canonical_model_id":"alibaba/qwen-flash"},"alibaba/deepseek-v4-flash":{"id":"alibaba/deepseek-v4-flash","name":"DeepSeek V4 Flash (Alibaba Cloud)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.2,"output":0.4,"cache_read":0.04},"canonical_model_id":"deepseek/deepseek-v4-flash"},"alibaba/qwen3.5-122b-a10b":{"id":"alibaba/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"alibaba/deepseek-v4.1-flash-1017":{"id":"alibaba/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Alibaba Cloud)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen3.8 Max (Alibaba Cloud)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"alibaba/glm-5.2":{"id":"alibaba/glm-5.2","name":"GLM-5.2 (Alibaba Cloud)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.28},"canonical_model_id":"zhipuai/glm-5.2"},"alibaba/qwen3.8-flash":{"id":"alibaba/qwen3.8-flash","name":"Qwen3.8 Flash (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"alibaba/qwen-plus":{"id":"alibaba/qwen-plus","name":"Qwen Plus (Alibaba Cloud)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32000},"cost":{"input":0.4,"output":1.2,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen-plus"},"alibaba/qwen35-397b-a17b":{"id":"alibaba/qwen35-397b-a17b","name":"Qwen3.5 397B A17B (Alibaba Cloud)","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"alibaba/qwen3.7-plus":{"id":"alibaba/qwen3.7-plus","name":"Qwen3.7 Plus (Alibaba Cloud)","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"alibaba/qwen-max":{"id":"alibaba/qwen-max","name":"Qwen Max (Alibaba Cloud)","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4},"canonical_model_id":"alibaba/qwen-max"},"alibaba/qwen3.5-27b":{"id":"alibaba/qwen3.5-27b","name":"Qwen3.5 27B (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"alibaba/qwen3.6-35b-a3b":{"id":"alibaba/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B (Alibaba Cloud)","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.375,"output":2.25},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"alibaba/qwen3.6-flash":{"id":"alibaba/qwen3.6-flash","name":"Qwen3.6 Flash (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.05,"cache_write":0.3125},"canonical_model_id":"alibaba/qwen3.6-flash"},"groq/gpt-oss-20b":{"id":"groq/gpt-oss-20b","name":"GPT OSS 20B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-oss-20b"},"groq/gpt-oss-120b":{"id":"groq/gpt-oss-120b","name":"GPT OSS 120B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.15,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"runpod/kimi-k3":{"id":"runpod/kimi-k3","name":"Kimi K3 (Runpod)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"aws-bedrock/claude-haiku-4-5-20251001":{"id":"aws-bedrock/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5 (2025-10-01) (AWS Bedrock)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"aws-bedrock/kimi-k3":{"id":"aws-bedrock/kimi-k3","name":"Kimi K3 (AWS Bedrock)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"moonshotai/kimi-k3"},"aws-bedrock/claude-opus-4-5-20251101":{"id":"aws-bedrock/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (AWS Bedrock)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"aws-bedrock/claude-haiku-4-5":{"id":"aws-bedrock/claude-haiku-4-5","name":"Claude Haiku 4.5 (AWS Bedrock)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"aws-bedrock/claude-sonnet-5-5":{"id":"aws-bedrock/claude-sonnet-5-5","name":"Claude Sonnet 5.5 (AWS Bedrock)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"aws-bedrock/claude-opus-5-5":{"id":"aws-bedrock/claude-opus-5-5","name":"Claude Opus 5.5 (AWS Bedrock)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"aws-bedrock/claude-sonnet-4-5-20250929":{"id":"aws-bedrock/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5 (2025-09-29) (AWS Bedrock)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"aws-bedrock/claude-fable-5":{"id":"aws-bedrock/claude-fable-5","name":"Claude Fable 5 (AWS Bedrock)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"aws-bedrock/claude-fable-5-1":{"id":"aws-bedrock/claude-fable-5-1","name":"Claude Fable 5.1 (AWS Bedrock)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"aws-bedrock/claude-opus-4-7":{"id":"aws-bedrock/claude-opus-4-7","name":"Claude Opus 4.7 (AWS Bedrock)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"aws-bedrock/claude-sonnet-4-5":{"id":"aws-bedrock/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (AWS Bedrock)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"aws-bedrock/grok-4-7":{"id":"aws-bedrock/grok-4-7","name":"Grok 4.7 (AWS Bedrock)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":524288},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"aws-bedrock/llama-3.1-70b-instruct":{"id":"aws-bedrock/llama-3.1-70b-instruct","name":"Llama 3.1 70B Instruct (AWS Bedrock)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"aws-bedrock/claude-opus-4-6":{"id":"aws-bedrock/claude-opus-4-6","name":"Claude Opus 4.6 (AWS Bedrock)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"aws-bedrock/claude-opus-5":{"id":"aws-bedrock/claude-opus-5","name":"Claude Opus 5 (AWS Bedrock)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"aws-bedrock/grok-4-3":{"id":"aws-bedrock/grok-4-3","name":"Grok 4.3 (AWS Bedrock)","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"aws-bedrock/claude-opus-4-8":{"id":"aws-bedrock/claude-opus-4-8","name":"Claude Opus 4.8 (AWS Bedrock)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"aws-bedrock/claude-sonnet-5":{"id":"aws-bedrock/claude-sonnet-5","name":"Claude Sonnet 5 (AWS Bedrock)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"aws-bedrock/llama-4-scout-17b-instruct":{"id":"aws-bedrock/llama-4-scout-17b-instruct","name":"Llama 4 Scout 17B Instruct (AWS Bedrock)","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"status":"deprecated","cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"aws-bedrock/grok-4-6":{"id":"aws-bedrock/grok-4-6","name":"Grok 4.6 (AWS Bedrock)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"aws-bedrock/claude-haiku-5-5":{"id":"aws-bedrock/claude-haiku-5-5","name":"Claude Haiku 5.5 (AWS Bedrock)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"aws-bedrock/claude-sonnet-4-6":{"id":"aws-bedrock/claude-sonnet-4-6","name":"Claude Sonnet 4.6 (AWS Bedrock)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"aws-bedrock/llama-4-maverick-17b-instruct":{"id":"aws-bedrock/llama-4-maverick-17b-instruct","name":"Llama 4 Maverick 17B Instruct (AWS Bedrock)","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"status":"deprecated","cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5 (MiniMax)","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7 (MiniMax)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax M2.5 Highspeed (MiniMax)","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.6,"output":2.4,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1 (MiniMax)","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.27,"output":1.1},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.1-lightning":{"id":"minimax/minimax-m2.1-lightning","name":"MiniMax M2.1 Lightning (MiniMax)","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":131072},"cost":{"input":0.12,"output":0.48}},"minimax/minimax-text-01":{"id":"minimax/minimax-text-01","name":"MiniMax Text 01 (MiniMax)","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2,"output":1.1}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3 (MiniMax)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax M3.1 Flash (MiniMax)","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2 (MiniMax)","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.2,"output":1,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 Highspeed (MiniMax)","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.6,"output":2.4,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"vertex-anthropic/claude-opus-4-5-20251101":{"id":"vertex-anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (Vertex AI (Anthropic))","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"vertex-anthropic/claude-haiku-4-5":{"id":"vertex-anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (Vertex AI (Anthropic))","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"vertex-anthropic/claude-opus-4-7":{"id":"vertex-anthropic/claude-opus-4-7","name":"Claude Opus 4.7 (Vertex AI (Anthropic))","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"vertex-anthropic/claude-sonnet-4-5":{"id":"vertex-anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (Vertex AI (Anthropic))","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"vertex-anthropic/claude-opus-4-6":{"id":"vertex-anthropic/claude-opus-4-6","name":"Claude Opus 4.6 (Vertex AI (Anthropic))","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"vertex-anthropic/claude-sonnet-5":{"id":"vertex-anthropic/claude-sonnet-5","name":"Claude Sonnet 5 (Vertex AI (Anthropic))","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"vertex-anthropic/claude-sonnet-4-6":{"id":"vertex-anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6 (Vertex AI (Anthropic))","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"google-ai-studio/gemini-3.1-flash-lite":{"id":"google-ai-studio/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite (Google AI Studio)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google-ai-studio/gemini-pro-latest":{"id":"google-ai-studio/gemini-pro-latest","name":"Gemini Pro Latest (Google AI Studio)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-27","last_updated":"2026-02-27","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2}},"google-ai-studio/gemini-3.6-flash":{"id":"google-ai-studio/gemini-3.6-flash","name":"Gemini 3.6 Flash (Google AI Studio)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.6-flash"},"google-ai-studio/gemini-3.5-flash-lite":{"id":"google-ai-studio/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite (Google AI Studio)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google-ai-studio/gemini-4-argon":{"id":"google-ai-studio/gemini-4-argon","name":"Gemini 4 Argon (Google AI Studio)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"google-ai-studio/gemini-3.1-pro-preview":{"id":"google-ai-studio/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview) (Google AI Studio)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google-ai-studio/gemini-3-flash-preview":{"id":"google-ai-studio/gemini-3-flash-preview","name":"Gemini 3 Flash (Preview) (Google AI Studio)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google-ai-studio/gemini-3.8-flash":{"id":"google-ai-studio/gemini-3.8-flash","name":"Gemini 3.8 Flash (Google AI Studio)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.8-flash"},"baidu/deepseek-v4.1-flash":{"id":"baidu/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Baidu)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"baidu/glm-5.3":{"id":"baidu/glm-5.3","name":"GLM-5.3 (Baidu)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"baidu/deepseek-v4-pro":{"id":"baidu/deepseek-v4-pro","name":"DeepSeek V4 Pro (Baidu)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.32,"output":3.96,"cache_read":0.042},"canonical_model_id":"deepseek/deepseek-v4-pro"},"baidu/glm-5.1":{"id":"baidu/glm-5.1","name":"GLM-5.1 (Baidu)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"baidu/deepseek-v4-flash":{"id":"baidu/deepseek-v4-flash","name":"DeepSeek V4 Flash (Baidu)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"baidu/kimi-k2.6":{"id":"baidu/kimi-k2.6","name":"Kimi K2.6 (Baidu)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"baidu/glm-5.2":{"id":"baidu/glm-5.2","name":"GLM-5.2 (Baidu)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"entrim/deepseek-v4.1-flash":{"id":"entrim/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (entrim)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"entrim/glm-5.3":{"id":"entrim/glm-5.3","name":"GLM-5.3 (entrim)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.5,"output":2,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-5.3"},"entrim/glm-5.3-flash":{"id":"entrim/glm-5.3-flash","name":"GLM-5.3 Flash (entrim)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.0175},"canonical_model_id":"zhipuai/glm-5.3-flash"},"azure-ai-foundry/grok-4-3":{"id":"azure-ai-foundry/grok-4-3","name":"Grok 4.3 (Azure AI Foundry)","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":20000,"output":8192},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"azure-ai-foundry/grok-4-1-fast-non-reasoning":{"id":"azure-ai-foundry/grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast Non-Reasoning (Azure AI Foundry)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5}},"azure-ai-foundry/grok-4-1-fast-reasoning":{"id":"azure-ai-foundry/grok-4-1-fast-reasoning","name":"Grok 4.1 Fast Reasoning (Azure AI Foundry)","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo V2.6 Flash (Xiaomi)","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo V2.6 Pro (Xiaomi)","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"scx-ai/qwen3-32b":{"id":"scx-ai/qwen3-32b","name":"Qwen3 32B (SCX.ai (Turbo))","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.36,"output":0.87},"canonical_model_id":"alibaba/qwen3-32b"},"scx-ai/minimax-m2.7":{"id":"scx-ai/minimax-m2.7","name":"MiniMax M2.7 (SCX.ai (Turbo))","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":196608},"cost":{"input":0.48,"output":1.79,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M2.7"},"scx-ai/gpt-oss-120b":{"id":"scx-ai/gpt-oss-120b","name":"GPT OSS 120B (SCX.ai (Turbo))","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.17,"output":0.55},"canonical_model_id":"openai/gpt-oss-120b"},"mistral/codestral-2508":{"id":"mistral/codestral-2508","name":"Codestral (Mistral AI)","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-2508"},"mistral/glm-5.3":{"id":"mistral/glm-5.3","name":"GLM-5.3 (Mistral AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3 (Mistral AI)","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/ministral-14b-2512":{"id":"mistral/ministral-14b-2512","name":"Ministral 14B (Mistral AI)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"mistral/ministral-14b-2512"},"mistral/devstral-2512":{"id":"mistral/devstral-2512","name":"Devstral 2 (Mistral AI)","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-2512"},"mistral/ministral-3b-2512":{"id":"mistral/ministral-3b-2512","name":"Ministral 3B (Mistral AI)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"mistral/ministral-3b-2512"},"mistral/mistral-medium-3-5":{"id":"mistral/mistral-medium-3-5","name":"Mistral Medium 3.5 (Mistral AI)","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistral/mistral-large-4":{"id":"mistral/mistral-large-4","name":"Mistral Large 4 (Mistral AI)","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistral/mistral-small-2506":{"id":"mistral/mistral-small-2506","name":"Mistral Small 3.2 (Mistral AI)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/mistral-small-2506"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large Latest (Mistral AI)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"cost":{"input":4,"output":12},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/mistral-small-2603":{"id":"mistral/mistral-small-2603","name":"Mistral Small 4 (Mistral AI)","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-2603"},"mistral/ministral-8b-2512":{"id":"mistral/ministral-8b-2512","name":"Ministral 8B (Mistral AI)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.15},"canonical_model_id":"mistral/ministral-8b-2512"},"openai/gpt-live-transcribe":{"id":"openai/gpt-live-transcribe","type":"transcription","name":"GPT Live Transcribe (OpenAI)","description":"Low-latency streaming speech-to-text model with keyword and language hints","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-live-transcribe"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra (OpenAI)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro (OpenAI)","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol (OpenAI)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna (OpenAI)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4 (OpenAI)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna (OpenAI)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex (OpenAI)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-transcribe":{"id":"openai/gpt-transcribe","type":"transcription","name":"GPT Transcribe (OpenAI)","description":"High-accuracy speech-to-text model for audio files and committed Realtime turns","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-transcribe"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o Mini (OpenAI)","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2 (OpenAI)","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini (OpenAI)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini (OpenAI)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra (OpenAI)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro (OpenAI)","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro (OpenAI)","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-4o-mini-transcribe":{"id":"openai/gpt-4o-mini-transcribe","name":"GPT-4o Mini Transcribe (OpenAI)","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"cost":{"input":1.25,"output":5,"input_audio":3}},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro (OpenAI)","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1 (OpenAI)","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-4o-transcribe":{"id":"openai/gpt-4o-transcribe","name":"GPT-4o Transcribe (OpenAI)","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"cost":{"input":2.5,"output":10,"input_audio":6}},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano (OpenAI)","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o (OpenAI)","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5 (OpenAI)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano (OpenAI)","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/o3":{"id":"openai/o3","name":"o3 (OpenAI)","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5 (OpenAI)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol (OpenAI)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol (OpenAI)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1 (OpenAI)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini (OpenAI)","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar (Perplexity)","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":130000,"output":4096},"cost":{"input":0.25,"output":2.5,"cache_write":0.25},"canonical_model_id":"perplexity/sonar"},"bytedance/seed-1-6-flash-250715":{"id":"bytedance/seed-1-6-flash-250715","name":"Seed 1.6 Flash (250715) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.07,"output":0.3,"cache_read":0.015}},"bytedance/seed-1-8-251228":{"id":"bytedance/seed-1-8-251228","name":"Seed 1.8 (251228) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/deepseek-v4-pro":{"id":"bytedance/deepseek-v4-pro","name":"DeepSeek V4 Pro (ByteDance)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"bytedance/seed-2-0-mini-260428":{"id":"bytedance/seed-2-0-mini-260428","name":"Seed 2.0 Mini (260428) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.4,"cache_read":0.02}},"bytedance/seed-2-0-pro-260328":{"id":"bytedance/seed-2-0-pro-260328","name":"Seed 2.0 Pro (260328) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"bytedance/deepseek-v4-flash":{"id":"bytedance/deepseek-v4-flash","name":"DeepSeek V4 Flash (ByteDance)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"bytedance/seed-2-0-lite-260428":{"id":"bytedance/seed-2-0-lite-260428","name":"Seed 2.0 Lite (260428) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/glm-4.7":{"id":"bytedance/glm-4.7","name":"GLM-4.7 (ByteDance)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"bytedance/glm-5.2":{"id":"bytedance/glm-5.2","name":"GLM-5.2 (ByteDance)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"bytedance/seed-1-6-250915":{"id":"bytedance/seed-1-6-250915","name":"Seed 1.6 (250915) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/gpt-oss-120b":{"id":"bytedance/gpt-oss-120b","name":"GPT OSS 120B (ByteDance)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.1,"output":0.5,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-120b"},"bytedance/seed-2-0-code-preview-260328":{"id":"bytedance/seed-2-0-code-preview-260328","name":"Seed 2.0 Code Preview (260328) (ByteDance)","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"bytedance/seed-1-6-250615":{"id":"bytedance/seed-1-6-250615","name":"Seed 1.6 (250615) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-06-25","last_updated":"2025-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/deepseek-v3.2":{"id":"bytedance/deepseek-v3.2","name":"DeepSeek V3.2 (ByteDance)","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.28,"output":0.42,"cache_read":0.056},"canonical_model_id":"deepseek/deepseek-v3.2"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1 (Meta)","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder 1.0 (Meta)","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2 (Meta)","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3 (Meta)","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3 (Moonshot AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.5":{"id":"moonshot/kimi-k2.5","name":"Kimi K2.5 (Moonshot AI)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code (Moonshot AI)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3.1-pro":{"id":"moonshot/kimi-k3.1-pro","name":"Kimi K3.1 Pro (Moonshot AI)","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6 (Moonshot AI)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k3.1-pro-highspeed":{"id":"moonshot/kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed (Moonshot AI)","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4}},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed (Moonshot AI)","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"canopywave/kimi-k3":{"id":"canopywave/kimi-k3","name":"Kimi K3 (CanopyWave)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"canopywave/deepseek-v4-pro":{"id":"canopywave/deepseek-v4-pro","name":"DeepSeek V4 Pro (CanopyWave)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.74,"output":3.48,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4-pro"},"canopywave/deepseek-v4-flash":{"id":"canopywave/deepseek-v4-flash","name":"DeepSeek V4 Flash (CanopyWave)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash"},"canopywave/kimi-k2.6":{"id":"canopywave/kimi-k2.6","name":"Kimi K2.6 (CanopyWave)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"canopywave/glm-5.2":{"id":"canopywave/glm-5.2","name":"GLM-5.2 (CanopyWave)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32768},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"runware/deepseek-v4.1-flash":{"id":"runware/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Runware)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"runware/glm-5.3":{"id":"runware/glm-5.3","name":"GLM-5.3 (Runware)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"runware/kimi-k3":{"id":"runware/kimi-k3","name":"Kimi K3 (Runware)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"runware/deepseek-v4-pro":{"id":"runware/deepseek-v4-pro","name":"DeepSeek V4 Pro (Runware)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.961,"output":1.922,"cache_read":0.079},"canonical_model_id":"deepseek/deepseek-v4-pro"},"runware/deepseek-v4-flash":{"id":"runware/deepseek-v4-flash","name":"DeepSeek V4 Flash (Runware)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.076,"output":0.153,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"runware/deepseek-v4.1-flash-1017":{"id":"runware/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (runware)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.6,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"runware/kimi-k2.6":{"id":"runware/kimi-k2.6","name":"Kimi K2.6 (Runware)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.6,"output":3.05,"cache_read":0.13},"canonical_model_id":"moonshotai/kimi-k2.6"},"runware/glm-5.2":{"id":"runware/glm-5.2","name":"GLM-5.2 (Runware)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":128000},"cost":{"input":0.8,"output":2.55,"cache_read":0.16},"canonical_model_id":"zhipuai/glm-5.2"},"runware/glm-5.3-flash":{"id":"runware/glm-5.3-flash","name":"GLM-5.3 Flash (Runware)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"runware/gpt-oss-120b":{"id":"runware/gpt-oss-120b","name":"GPT OSS 120B (Runware)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.032,"output":0.14,"cache_read":0.032},"canonical_model_id":"openai/gpt-oss-120b"},"runware/gemma-4-31b-it":{"id":"runware/gemma-4-31b-it","name":"Gemma 4 31B IT (Runware)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.102,"output":0.297,"cache_read":0.012},"canonical_model_id":"google/gemma-4-31b-it"},"gonka24/minimax-m2.7":{"id":"gonka24/minimax-m2.7","name":"MiniMax M2.7 (Gonka24)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.08,"output":0.32,"cache_read":0.017},"canonical_model_id":"minimax/MiniMax-M2.7"},"azure-anthropic/claude-sonnet-5-5":{"id":"azure-anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5 (Azure Anthropic)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"azure-anthropic/claude-opus-5-5":{"id":"azure-anthropic/claude-opus-5-5","name":"Claude Opus 5.5 (Azure Anthropic)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"azure-anthropic/claude-fable-5":{"id":"azure-anthropic/claude-fable-5","name":"Claude Fable 5 (Azure Anthropic)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"azure-anthropic/claude-fable-5-1":{"id":"azure-anthropic/claude-fable-5-1","name":"Claude Fable 5.1 (Azure Anthropic)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"azure-anthropic/claude-opus-4-7":{"id":"azure-anthropic/claude-opus-4-7","name":"Claude Opus 4.7 (Azure Anthropic)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"azure-anthropic/claude-opus-4-6":{"id":"azure-anthropic/claude-opus-4-6","name":"Claude Opus 4.6 (Azure Anthropic)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"azure-anthropic/claude-opus-5":{"id":"azure-anthropic/claude-opus-5","name":"Claude Opus 5 (Azure Anthropic)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"azure-anthropic/claude-opus-4-8":{"id":"azure-anthropic/claude-opus-4-8","name":"Claude Opus 4.8 (Azure Anthropic)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"azure-anthropic/claude-sonnet-5":{"id":"azure-anthropic/claude-sonnet-5","name":"Claude Sonnet 5 (Azure Anthropic)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"embercloud/glm-5":{"id":"embercloud/glm-5","name":"GLM-5 (EmberCloud)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131000},"cost":{"input":0.72,"output":2.3,"cache_read":0.144},"canonical_model_id":"zhipuai/glm-5"},"embercloud/glm-5.1":{"id":"embercloud/glm-5.1","name":"GLM-5.1 (EmberCloud)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131000},"cost":{"input":0.931,"output":2.93,"cache_read":0.173},"canonical_model_id":"zhipuai/glm-5.1"},"embercloud/kimi-k2.5":{"id":"embercloud/kimi-k2.5","name":"Kimi K2.5 (EmberCloud)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.405,"output":1.98,"cache_read":0.225},"canonical_model_id":"moonshotai/kimi-k2.5"},"embercloud/glm-4.7-flash":{"id":"embercloud/glm-4.7-flash","name":"GLM-4.7 Flash (EmberCloud)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131000},"cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"embercloud/glm-4.7":{"id":"embercloud/glm-4.7","name":"GLM-4.7 (EmberCloud)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131000},"cost":{"input":0.38,"output":1.98,"cache_read":0.19},"canonical_model_id":"zhipuai/glm-4.7"},"embercloud/glm-4.5":{"id":"embercloud/glm-4.5","name":"GLM-4.5 (EmberCloud)","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":96000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"embercloud/glm-5.2":{"id":"embercloud/glm-5.2","name":"GLM-5.2 (EmberCloud)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131000},"cost":{"input":1.26,"output":3.96,"cache_read":0.234},"canonical_model_id":"zhipuai/glm-5.2"},"embercloud/glm-4.5-air":{"id":"embercloud/glm-4.5-air","name":"GLM-4.5 Air (EmberCloud)","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":96000},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"},"embercloud/qwen3-coder-next":{"id":"embercloud/qwen3-coder-next","name":"Qwen3 Coder Next (EmberCloud)","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.108,"output":0.675,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3-coder-next"},"tencent/glm-5":{"id":"tencent/glm-5","name":"GLM-5 (Tencent Cloud)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"tencent/glm-5-turbo":{"id":"tencent/glm-5-turbo","name":"GLM-5 Turbo (Tencent Cloud)","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"tencent/kimi-k3":{"id":"tencent/kimi-k3","name":"Kimi K3 (Tencent Cloud)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"tencent/deepseek-v4-pro":{"id":"tencent/deepseek-v4-pro","name":"DeepSeek V4 Pro (Tencent Cloud)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.435,"output":0.87,"cache_read":0.00363},"canonical_model_id":"deepseek/deepseek-v4-pro"},"tencent/glm-5.1":{"id":"tencent/glm-5.1","name":"GLM-5.1 (Tencent Cloud)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3 (Tencent Cloud)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"tencent/kimi-k2.7-code":{"id":"tencent/kimi-k2.7-code","name":"Kimi K2.7 Code (Tencent Cloud)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"tencent/minimax-m2.7":{"id":"tencent/minimax-m2.7","name":"MiniMax M2.7 (Tencent Cloud)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"tencent/deepseek-v4-flash":{"id":"tencent/deepseek-v4-flash","name":"DeepSeek V4 Flash (Tencent Cloud)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"tencent/hy-mt2-plus":{"id":"tencent/hy-mt2-plus","name":"Hy-MT2 Plus (Tencent Cloud)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295}},"tencent/kimi-k2.6":{"id":"tencent/kimi-k2.6","name":"Kimi K2.6 (Tencent Cloud)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.858,"output":3.566,"cache_read":0.145},"canonical_model_id":"moonshotai/kimi-k2.6"},"tencent/minimax-m3":{"id":"tencent/minimax-m3","name":"MiniMax M3 (Tencent Cloud)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"tencent/glm-5.2":{"id":"tencent/glm-5.2","name":"GLM-5.2 (Tencent Cloud)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"tencent/glm-5v-turbo":{"id":"tencent/glm-5v-turbo","name":"GLM-5V Turbo (Tencent Cloud)","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 Preview (Tencent Cloud)","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/kimi-k2.7-code-highspeed":{"id":"tencent/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed (Tencent Cloud)","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"anthropic/claude-haiku-4-5-20251001":{"id":"anthropic/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5 (2025-10-01) (Anthropic)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (Anthropic)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (Anthropic)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5 (Anthropic)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5 (Anthropic)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-sonnet-4-5-20250929":{"id":"anthropic/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5 (2025-09-29) (Anthropic)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5 (Anthropic)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1 (Anthropic)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7 (Anthropic)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (Anthropic)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6 (Anthropic)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5 (Anthropic)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8 (Anthropic)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5 (Anthropic)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5 (Anthropic)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6 (Anthropic)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"consensusprotocol/deepseek-v4.1-flash":{"id":"consensusprotocol/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Consensus Protocol)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.55,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"consensusprotocol/mimo-v2.6-flash":{"id":"consensusprotocol/mimo-v2.6-flash","name":"MiMo V2.6 Flash (Consensus Protocol)","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.12,"output":0.25,"cache_read":0.0025},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"consensusprotocol/gpt-oss-20b":{"id":"consensusprotocol/gpt-oss-20b","name":"GPT OSS 20B (Consensus Protocol)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.04,"output":0.19,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-20b"},"consensusprotocol/deepseek-v4-flash":{"id":"consensusprotocol/deepseek-v4-flash","name":"DeepSeek V4 Flash (Consensus Protocol)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.1,"output":0.2,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4-flash"},"consensusprotocol/minimax-m3":{"id":"consensusprotocol/minimax-m3","name":"MiniMax M3 (Consensus Protocol)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"consensusprotocol/glm-5.3-flash":{"id":"consensusprotocol/glm-5.3-flash","name":"GLM-5.3 Flash (Consensus Protocol)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.25,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-5.3-flash"},"consensusprotocol/Qwen3.8-27B":{"id":"consensusprotocol/Qwen3.8-27B","name":"Qwen3.8 27B (Consensus Protocol)","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":32768},"cost":{"input":0.08,"output":0.35,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"consensusprotocol/gemma-4-31b-it":{"id":"consensusprotocol/gemma-4-31b-it","name":"Gemma 4 31B IT (Consensus Protocol)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.25,"cache_read":0.01},"canonical_model_id":"google/gemma-4-31b-it"},"greenference/qwen3-14b":{"id":"greenference/qwen3-14b","name":"Qwen3 14B (greenference)","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":0.076,"output":0.228}},"greenference/qwen3.6-27b":{"id":"greenference/qwen3.6-27b","name":"Qwen3.6 27B (greenference)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":8192},"cost":{"input":0.15,"output":1.35},"canonical_model_id":"alibaba/qwen3.6-27b"},"greenference/glm-5.3-flash":{"id":"greenference/glm-5.3-flash","name":"GLM-5.3 Flash (greenference)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.49,"cache_read":0.028},"canonical_model_id":"zhipuai/glm-5.3-flash"},"greenference/qwen3-30b-a3b":{"id":"greenference/qwen3-30b-a3b","name":"Qwen3 30B A3B (greenference)","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.095,"output":0.285},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"google-vertex/gemini-3.1-flash-lite":{"id":"google-vertex/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite (Google Vertex AI)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google-vertex/gemini-3.6-flash":{"id":"google-vertex/gemini-3.6-flash","name":"Gemini 3.6 Flash (Google Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.6-flash"},"google-vertex/gemini-3.5-flash-lite":{"id":"google-vertex/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite (Google Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google-vertex/gemini-4-argon":{"id":"google-vertex/gemini-4-argon","name":"Gemini 4 Argon (Google Vertex AI)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"google-vertex/gemini-3.1-pro-preview":{"id":"google-vertex/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview) (Google Vertex AI)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google-vertex/gemini-3.7-flash":{"id":"google-vertex/gemini-3.7-flash","name":"Gemini 3.7 Flash (Google Vertex AI)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.7-flash"},"google-vertex/gemini-3-flash-preview":{"id":"google-vertex/gemini-3-flash-preview","name":"Gemini 3 Flash (Preview) (Google Vertex AI)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google-vertex/gemini-3.8-flash":{"id":"google-vertex/gemini-3.8-flash","name":"Gemini 3.8 Flash (Google Vertex AI)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.8-flash"},"google-vertex/gemini-3.5-flash":{"id":"google-vertex/gemini-3.5-flash","name":"Gemini 3.5 Flash (Google Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.08333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"quartz/gemini-3.1-pro-preview":{"id":"quartz/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview) (Quartz)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"ranoai/deepseek-v4-flash":{"id":"ranoai/deepseek-v4-flash","name":"DeepSeek V4 Flash (RanoAI)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"scx-ai-gp/deepseek-v4.1-flash":{"id":"scx-ai-gp/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (SCX.ai)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"scx-ai-gp/glm-5.3":{"id":"scx-ai-gp/glm-5.3","name":"GLM-5.3 (SCX.ai)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.924,"output":2.904,"cache_read":0.1716},"canonical_model_id":"zhipuai/glm-5.3"},"scx-ai-gp/kimi-k3":{"id":"scx-ai-gp/kimi-k3","name":"Kimi K3 (SCX.ai)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3.135,"output":15.675,"cache_read":0.319},"canonical_model_id":"moonshotai/kimi-k3"},"scx-ai-gp/glm-5.2-fast":{"id":"scx-ai-gp/glm-5.2-fast","name":"GLM-5.2 Turbo (SCX.ai)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.156,"output":6.776,"cache_read":0.429},"canonical_model_id":"zhipuai/glm-5.2"},"scx-ai-gp/kimi-k2.7-code":{"id":"scx-ai-gp/kimi-k2.7-code","name":"Kimi K2.7 Code (SCX.ai)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.737,"output":3.08,"cache_read":0.143},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"scx-ai-gp/glm-5.3-fast":{"id":"scx-ai-gp/glm-5.3-fast","name":"GLM 5.3 Fast (SCX.ai)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":2.8,"output":8.8,"cache_read":0.56},"canonical_model_id":"zhipuai/glm-5.3"},"scx-ai-gp/qwen3.8-max":{"id":"scx-ai-gp/qwen3.8-max","name":"Qwen3.8 Max (SCX.ai)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.54,"output":4.631,"cache_read":0.198,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"scx-ai-gp/glm-5.2":{"id":"scx-ai-gp/glm-5.2","name":"GLM-5.2 (SCX.ai)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.605,"output":2.123,"cache_read":0.154},"canonical_model_id":"zhipuai/glm-5.2"},"scx-ai-gp/glm-5.3-flash":{"id":"scx-ai-gp/glm-5.3-flash","name":"GLM-5.3 Flash (SCX.ai)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.088,"output":0.25,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-5.3-flash"},"fireworks/deepseek-v4.1-flash":{"id":"fireworks/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Fireworks AI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"fireworks/glm-5.3":{"id":"fireworks/glm-5.3","name":"GLM-5.3 (Fireworks AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"fireworks/kimi-k3":{"id":"fireworks/kimi-k3","name":"Kimi K3 (Fireworks AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1040384,"output":1040384},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"fireworks/glm-5.3-flash":{"id":"fireworks/glm-5.3-flash","name":"GLM-5.3 Flash (Fireworks AI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"fireworks/kimi-k3-fast":{"id":"fireworks/kimi-k3-fast","name":"Kimi K3 Fast (Fireworks AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1040384,"output":1040384},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"novita/glm-5":{"id":"novita/glm-5","name":"GLM-5 (NovitaAI)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"novita/qwen3.8-27b":{"id":"novita/qwen3.8-27b","name":"Qwen3.8 27B (NovitaAI)","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.42,"output":3,"cache_read":0.085},"canonical_model_id":"alibaba/qwen3.8-27b"},"novita/qwen3-vl-235b-a22b-thinking":{"id":"novita/qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking (NovitaAI)","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"novita/ernie-4.5-vl-424b-a47b":{"id":"novita/ernie-4.5-vl-424b-a47b","name":"ERNIE 4.5 VL 424B A47B (NovitaAI)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":123000,"output":16000},"cost":{"input":0.42,"output":1.25}},"novita/deepseek-v4.1-flash":{"id":"novita/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (NovitaAI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"novita/glm-5.3":{"id":"novita/glm-5.3","name":"GLM-5.3 (NovitaAI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"novita/llama-3.3-70b-instruct":{"id":"novita/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct (NovitaAI)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":120000},"cost":{"input":0.135,"output":0.4},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"novita/kimi-k3":{"id":"novita/kimi-k3","name":"Kimi K3 (NovitaAI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"novita/minimax-m2.5":{"id":"novita/minimax-m2.5","name":"MiniMax M2.5 (NovitaAI)","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"novita/qwen3-vl-235b-a22b-instruct":{"id":"novita/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct (NovitaAI)","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"novita/qwen3-coder-30b-a3b-instruct":{"id":"novita/qwen3-coder-30b-a3b-instruct","name":"Qwen3 Coder 30B A3B Instruct (NovitaAI)","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":160000,"output":32768},"cost":{"input":0.07,"output":0.27},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"novita/glm-5.1":{"id":"novita/glm-5.1","name":"GLM-5.1 (NovitaAI)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"novita/kimi-k2":{"id":"novita/kimi-k2","name":"Kimi K2 (NovitaAI)","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":0.57,"output":2.3}},"novita/glm-4.5v":{"id":"novita/glm-4.5v","name":"GLM-4.5V (NovitaAI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"novita/qwen3.7-max":{"id":"novita/qwen3.7-max","name":"Qwen3.7 Max (NovitaAI)","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":3.75,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.7-max"},"novita/hy3":{"id":"novita/hy3","name":"Hy3 (NovitaAI)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"novita/qwen3-235b-a22b-instruct-2507":{"id":"novita/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507 (NovitaAI)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.09,"output":0.58},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"novita/kimi-k2.7-code":{"id":"novita/kimi-k2.7-code","name":"Kimi K2.7 Code (NovitaAI)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"novita/minimax-m2.7":{"id":"novita/minimax-m2.7","name":"MiniMax M2.7 (NovitaAI)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"novita/gemma-4-26b-a4b-it":{"id":"novita/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT (NovitaAI)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"novita/step-3.7-flash":{"id":"novita/step-3.7-flash","name":"Step 3.7 Flash (NovitaAI)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"novita/qwen3-max":{"id":"novita/qwen3-max","name":"Qwen3 Max (NovitaAI)","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.845,"output":3.38},"canonical_model_id":"alibaba/qwen3-max"},"novita/qwen3-235b-a22b-thinking-2507":{"id":"novita/qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22B Thinking 2507 (NovitaAI)","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":3},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"novita/deepseek-v4-flash":{"id":"novita/deepseek-v4-flash","name":"DeepSeek V4 Flash (NovitaAI)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"novita/glm-4.7":{"id":"novita/glm-4.7","name":"GLM-4.7 (NovitaAI)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"novita/qwen3.5-122b-a10b":{"id":"novita/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B (NovitaAI)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"novita/deepseek-v4.1-flash-1017":{"id":"novita/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (NovitaAI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"novita/minimax-m2.1":{"id":"novita/minimax-m2.1","name":"MiniMax M2.1 (NovitaAI)","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"novita/llama-3.2-3b-instruct":{"id":"novita/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct (NovitaAI)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-09-18","last_updated":"2024-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":32000},"cost":{"input":0.03,"output":0.05}},"novita/qwen3-235b-a22b-fp8":{"id":"novita/qwen3-235b-a22b-fp8","name":"Qwen3 235B A22B FP8 (NovitaAI)","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"novita/qwen3.8-max":{"id":"novita/qwen3.8-max","name":"Qwen3.8 Max (NovitaAI)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"novita/llama-3-70b-instruct":{"id":"novita/llama-3-70b-instruct","name":"Llama 3 70B Instruct (NovitaAI)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-04-18","last_updated":"2024-04-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8000},"cost":{"input":0.51,"output":0.74}},"novita/kimi-k2.6":{"id":"novita/kimi-k2.6","name":"Kimi K2.6 (NovitaAI)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.8,"output":3.4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"novita/ling-3.0-flash":{"id":"novita/ling-3.0-flash","name":"InclusionAI Ling 3.0 Flash (NovitaAI)","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"novita/qwen3.8-2.4t-a95b":{"id":"novita/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B (NovitaAI)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"novita/glm-5.2":{"id":"novita/glm-5.2","name":"GLM-5.2 (NovitaAI)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"novita/qwen3.8-flash":{"id":"novita/qwen3.8-flash","name":"Qwen3.8 Flash (NovitaAI)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"novita/glm-5.3-flash":{"id":"novita/glm-5.3-flash","name":"GLM-5.3 Flash (NovitaAI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"novita/qwen35-397b-a17b":{"id":"novita/qwen35-397b-a17b","name":"Qwen3.5 397B A17B (NovitaAI)","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"novita/glm-4.6":{"id":"novita/glm-4.6","name":"GLM-4.6 (NovitaAI)","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"novita/glm-4.6v":{"id":"novita/glm-4.6v","name":"GLM-4.6V (NovitaAI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16000},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"novita/qwen3-vl-30b-a3b-instruct":{"id":"novita/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct (NovitaAI)","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"novita/gemma-4-31b-it":{"id":"novita/gemma-4-31b-it","name":"Gemma 4 31B IT (NovitaAI)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"novita/qwen3-coder-480b-a35b-instruct":{"id":"novita/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct (NovitaAI)","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"novita/llama-4-scout-17b-instruct":{"id":"novita/llama-4-scout-17b-instruct","name":"Llama 4 Scout 17B Instruct (NovitaAI)","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.18,"output":0.59},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"novita/qwen3-next-80b-a3b-instruct":{"id":"novita/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct (NovitaAI)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"novita/llama-4-maverick-17b-instruct":{"id":"novita/llama-4-maverick-17b-instruct","name":"Llama 4 Maverick 17B Instruct (NovitaAI)","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"novita/qwen3.6-35b-a3b":{"id":"novita/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B (NovitaAI)","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"novita/deepseek-v3.2":{"id":"novita/deepseek-v3.2","name":"DeepSeek V3.2 (NovitaAI)","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.269,"output":0.4,"cache_read":0.1345},"canonical_model_id":"deepseek/deepseek-v3.2"},"meta-contributor/muse-spark-1.3-contributor":{"id":"meta-contributor/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor (Meta Contributor)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta-contributor/muse-spark-1.2-contributor":{"id":"meta-contributor/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor (Meta Contributor)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"deepinfra/deepseek-v4.1-flash":{"id":"deepinfra/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (DeepInfra)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepinfra/qwen3-vl-235b-a22b-instruct":{"id":"deepinfra/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct (DeepInfra)","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"deepinfra/deepseek-v4-pro":{"id":"deepinfra/deepseek-v4-pro","name":"DeepSeek V4 Pro (DeepInfra)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepinfra/mimo-v2.6-flash":{"id":"deepinfra/mimo-v2.6-flash","name":"MiMo V2.6 Flash (DeepInfra)","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"deepinfra/granite-4.2-8b":{"id":"deepinfra/granite-4.2-8b","name":"Granite 4.2 8B (DeepInfra)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"deepinfra/gemma-4-26b-a4b-it":{"id":"deepinfra/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT (DeepInfra)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"deepinfra/qwen3.5-9b":{"id":"deepinfra/qwen3.5-9b","name":"Qwen3.5 9B (DeepInfra)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"deepinfra/nemotron-3-ultra-550b":{"id":"deepinfra/nemotron-3-ultra-550b","name":"Nemotron 3 Ultra 550B (DeepInfra)","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.2,"cache_read":0.1}},"deepinfra/mimo-v2.6-pro":{"id":"deepinfra/mimo-v2.6-pro","name":"MiMo V2.6 Pro (DeepInfra)","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"deepinfra/deepseek-v4-flash":{"id":"deepinfra/deepseek-v4-flash","name":"DeepSeek V4 Flash (DeepInfra)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.08,"output":0.18,"cache_read":0.016},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepinfra/deepseek-v4.1-flash-1017":{"id":"deepinfra/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (DeepInfra)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepinfra/muse-glimmer-30b":{"id":"deepinfra/muse-glimmer-30b","name":"Muse Glimmer 30B (DeepInfra)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"deepinfra/ling-3.0-flash-vl":{"id":"deepinfra/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL (DeepInfra)","description":"Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"deepinfra/ling-3.0-flash":{"id":"deepinfra/ling-3.0-flash","name":"InclusionAI Ling 3.0 Flash (DeepInfra)","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"deepinfra/qwen3-vl-30b-a3b-instruct":{"id":"deepinfra/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct (DeepInfra)","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"deepinfra/gemma-4-31b-it":{"id":"deepinfra/gemma-4-31b-it","name":"Gemma 4 31B IT (DeepInfra)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"deepinfra/deepseek-v3.2":{"id":"deepinfra/deepseek-v3.2","name":"DeepSeek V3.2 (DeepInfra)","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":160000,"output":65536},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"}}},"greenpt":{"id":"greenpt","env":["GREENPT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.greenpt.ai/v1","name":"GreenPT","doc":"https://docs.greenpt.ai","models":{"gemma-3-27b-it":{"id":"gemma-3-27b-it","name":"Gemma 3 27B","description":"Google Gemma 3 multimodal model for chat, reasoning, and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":8192},"status":"deprecated","cost":{"input":0.342,"output":0.684},"canonical_model_id":"google/gemma-3-27b-it"},"glm-5.2-honey-ultra":{"id":"glm-5.2-honey-ultra","name":"GLM-5.2 Honey Ultra","description":"glm-5.2 carrying a built-in ruleset that compresses both generated code and prose. The most aggressive tier, close to answer-only. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.255552,"output":1.27776,"cache_read":0.0127776},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.27754,"output":5.11016,"cache_read":0.319385},"canonical_model_id":"zhipuai/glm-5.3"},"green-r-raw":{"id":"green-r-raw","name":"Green R Raw","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.399,"output":1.083},"canonical_model_id":"openai/gpt-oss-120b"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":16384},"cost":{"input":1.254,"output":1.254},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3.762,"output":18.81,"cache_read":0.9405},"canonical_model_id":"moonshotai/kimi-k3"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.49346,"output":2.4673,"cache_read":0.123365},"canonical_model_id":"minimax/MiniMax-M2.5"},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":0.285,"output":1.083},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"status":"deprecated","cost":{"input":1.756,"output":5.518},"canonical_model_id":"zhipuai/glm-5.1"},"glm-5.2-ponytail-lite":{"id":"glm-5.2-ponytail-lite","name":"GLM-5.2 Ponytail Lite","description":"glm-5.2 carrying a built-in ruleset that compresses generated code, preferring platform features over custom code. The gentlest tier: it cuts filler only and keeps the explanation intact. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2-caveman-ultra":{"id":"glm-5.2-caveman-ultra","name":"GLM-5.2 Caveman Ultra","description":"glm-5.2 carrying a built-in ruleset that compresses prose, keeping code and technical detail verbatim. The most aggressive tier, close to answer-only. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2-ponytail-ultra":{"id":"glm-5.2-ponytail-ultra","name":"GLM-5.2 Ponytail Ultra","description":"glm-5.2 carrying a built-in ruleset that compresses generated code, preferring platform features over custom code. The most aggressive tier, close to answer-only. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen3 235B MoE instruct model for long-context multilingual chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":1.026,"output":3.078},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"pixtral-12b-2409":{"id":"pixtral-12b-2409","name":"Pixtral 12B","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-09-11","last_updated":"2024-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.285,"output":0.285},"canonical_model_id":"mistral/pixtral-12b"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.863555,"output":4.317775,"cache_read":0.222057},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"devstral-2-123b-instruct-2512":{"id":"devstral-2-123b-instruct-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":16384},"status":"deprecated","cost":{"input":0.57,"output":2.736},"canonical_model_id":"mistral/devstral-2512"},"green-l-raw":{"id":"green-l-raw","name":"Green L Raw","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.285,"output":0.912},"canonical_model_id":"mistral/mistral-small-2506"},"kimi-k2.6-fast":{"id":"kimi-k2.6-fast","name":"Kimi K2.6 Fast","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":1.655,"output":8.778},"canonical_model_id":"moonshotai/kimi-k2.6"},"glm-5.2-honey":{"id":"glm-5.2-honey","name":"GLM-5.2 Honey","description":"glm-5.2 carrying a built-in ruleset that compresses both generated code and prose. The middle tier, and the ruleset as its authors wrote it. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.228,"output":0.456},"canonical_model_id":"mistral/mistral-small-2506"},"mistral-medium-3.5-128b":{"id":"mistral-medium-3.5-128b","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":2.052,"output":10.26},"canonical_model_id":"mistral/mistral-medium-2604"},"glm-5.2-caveman":{"id":"glm-5.2-caveman","name":"GLM-5.2 Caveman","description":"glm-5.2 carrying a built-in ruleset that compresses prose, keeping code and technical detail verbatim. The middle tier, and the ruleset as its authors wrote it. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.798,"output":4.959},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.1596,"output":0.399,"cache_read":0.0456},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.7524,"output":4.275,"cache_read":0.2508},"canonical_model_id":"moonshotai/kimi-k2.6"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"green-s":{"id":"green-s","type":"transcription","name":"Green S","description":"GreenPT speech-to-text model for pre-recorded and live transcription","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":8192},"provider":{"api":"https://api.greenpt.ai"}},"glm-5.2-caveman-lite":{"id":"glm-5.2-caveman-lite","name":"GLM-5.2 Caveman Lite","description":"glm-5.2 carrying a built-in ruleset that compresses prose, keeping code and technical detail verbatim. The gentlest tier: it cuts filler only and keeps the explanation intact. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.127754,"output":0.511016,"cache_read":0.0255508},"canonical_model_id":"zhipuai/glm-5.3-flash"},"green-r":{"id":"green-r","name":"Green R","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.399,"output":1.083},"canonical_model_id":"openai/gpt-oss-120b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.228,"output":0.798},"canonical_model_id":"openai/gpt-oss-120b"},"green-l":{"id":"green-l","name":"Green L","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.285,"output":0.912},"canonical_model_id":"mistral/mistral-small-2506"},"glm-5.2-ponytail":{"id":"glm-5.2-ponytail","name":"GLM-5.2 Ponytail","description":"glm-5.2 carrying a built-in ruleset that compresses generated code, preferring platform features over custom code. The middle tier, and the ruleset as its authors wrote it. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2-honey-lite":{"id":"glm-5.2-honey-lite","name":"GLM-5.2 Honey Lite","description":"glm-5.2 carrying a built-in ruleset that compresses both generated code and prose. The gentlest tier: it cuts filler only and keeps the explanation intact. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"gemma4":{"id":"gemma4","name":"gemma4","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.57,"output":1.71},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"holo2-30b-a3b":{"id":"holo2-30b-a3b","name":"Holo2 30B A3B","description":"H Company Holo2 vision model for GUI navigation and computer-use agents","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11","last_updated":"2025-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":22016,"output":16384},"status":"deprecated","cost":{"input":0.399,"output":0.969}},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.342,"output":2.052},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"voxtral-small-24b-2507":{"id":"voxtral-small-24b-2507","name":"Voxtral Small 24B","description":"Mistral Voxtral audio-understanding model for speech and transcription tasks","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.228,"output":0.513},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"green-s-pro":{"id":"green-s-pro","type":"transcription","name":"Green S Pro","description":"GreenPT advanced speech-to-text model with multilingual transcription support","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-02","last_updated":"2025-02","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":8192},"provider":{"api":"https://api.greenpt.ai"}}}},"umans-ai":{"id":"umans-ai","env":["UMANS_AI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.code.umans.ai/v1","name":"Umans AI","doc":"https://app.umans.ai/offers/code/docs/orgs","models":{"umans-kimi-k3":{"id":"umans-kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"umans-coder":{"id":"umans-coder","name":"Umans Coder","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-glm-5.3-flash":{"id":"umans-glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-glm-5.3":{"id":"umans-glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"umans-deepseek-v4-flash-0731":{"id":"umans-deepseek-v4-flash-0731","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"umans-deepseek-v4.1-flash-1017":{"id":"umans-deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0.15,"output":0.6,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"umans-flash":{"id":"umans-flash","name":"Umans Flash","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"umans-deepseek-v4.1-flash":{"id":"umans-deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0.15,"output":0.6,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"}}},"scx-ai":{"id":"scx-ai","env":["SCX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.scx.ai/v1","name":"SCX.ai","doc":"https://platform.scx.ai/docs","models":{"Qwen3.8-Max":{"id":"Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":983616,"output":131072},"cost":{"input":1.815,"output":5.4461,"cache_read":0.17,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":196608},"cost":{"input":0.48,"output":1.79,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M2.7"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.17,"output":0.55},"canonical_model_id":"openai/gpt-oss-120b"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.55,"output":1.784,"cache_read":0.111},"canonical_model_id":"zhipuai/glm-5.2"}}},"stepfun":{"id":"stepfun","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.com/v1","name":"StepFun (China)","doc":"https://platform.stepfun.com/docs/zh/overview/concept","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.20175,"output":1.21052,"cache_read":0.04035},"canonical_model_id":"stepfun/step-3.7-flash"},"step-tts-2":{"id":"step-tts-2","type":"speech","name":"Step TTS 2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-01","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.04613,"output":2.98895,"cache_read":0.10461},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.10461,"output":0.31384,"cache_read":0.02092},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.10461,"output":0.31384,"cache_read":0.02092},"canonical_model_id":"stepfun/step-3.5-flash"},"stepaudio-2.5-asr":{"id":"stepaudio-2.5-asr","type":"transcription","name":"StepAudio 2.5 ASR","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"step","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-24","last_updated":"2026-07-02","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}},"stepaudio-2.5-tts":{"id":"stepaudio-2.5-tts","type":"speech","name":"StepAudio 2.5 TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}}}},"mistral":{"id":"mistral","env":["MISTRAL_API_KEY"],"npm":"@ai-sdk/mistral","name":"Mistral","doc":"https://docs.mistral.ai/getting-started/models/","models":{"codestral-2508":{"id":"codestral-2508","name":"Codestral 25.08","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"mistral-medium-latest":{"id":"mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-latest"},"glm-5-2":{"id":"glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05}},"voxtral-small-latest":{"id":"voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.1,"output":0.4}},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"Ministral 3 14B","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"devstral-2512":{"id":"devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"mistral-medium-2604":{"id":"mistral-medium-2604","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15}},"devstral-latest":{"id":"devstral-latest","name":"Devstral 2","description":"Legacy model retained for compatibility with older integrations","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"mistral-embed":{"id":"mistral-embed","type":"embedding","name":"Mistral Embed","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":3072},"cost":{"input":0.1,"output":0}},"ministral-3b-2512":{"id":"ministral-3b-2512","name":"Ministral 3 3B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":262144},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3-3b-instruct-2512"},"codestral-latest":{"id":"codestral-latest","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.3,"output":0.9,"cache_read":0.03}},"mistral-small-latest":{"id":"mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015}},"magistral-medium-latest":{"id":"magistral-medium-latest","name":"Magistral Medium (latest)","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-03-17","last_updated":"2025-03-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":2,"output":5}},"mistral-medium-2505":{"id":"mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":2}},"voxtral-mini-tts-latest":{"id":"voxtral-mini-tts-latest","type":"speech","name":"Voxtral Mini TTS (latest)","description":"Multilingual text-to-speech model with zero-shot voice cloning","family":"voxtral","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-23","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["audio"]},"open_weights":true,"limit":{"context":0,"output":0}},"voxtral-mini-latest":{"id":"voxtral-mini-latest","type":"transcription","name":"Voxtral Mini (latest)","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"ministral-8b-latest":{"id":"ministral-8b-latest","name":"Ministral 8B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-01","last_updated":"2024-10-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.15,"output":0.15,"cache_read":0.015}},"labs-leanstral-1-5-1":{"id":"labs-leanstral-1-5-1","name":"Leanstral 1.5","description":"An updated Lean 4 formal proof engineering model optimised for automated theorem proving and autoformalization. 119B total parameters, 6.5B active.","family":"leanstral","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":0,"output":0}},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"status":"beta","cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistral-small-2506":{"id":"mistral-small-2506","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.1,"output":0.3}},"zai-glm-5-3-flash":{"id":"zai-glm-5-3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"voxtral-small-2507":{"id":"voxtral-small-2507","name":"Voxtral Small","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.1,"output":0.4}},"open-mistral-nemo":{"id":"open-mistral-nemo","name":"Open Mistral Nemo","description":"Legacy model retained for compatibility with older integrations","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.15,"output":0.15}},"mistral-large-latest":{"id":"mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05}},"pixtral-12b":{"id":"pixtral-12b","name":"Pixtral 12B","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-09-01","last_updated":"2024-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.15,"output":0.15}},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015}},"labs-devstral-small-2512":{"id":"labs-devstral-small-2512","name":"Devstral Small 2","description":"Legacy model retained for compatibility with older integrations","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0}},"zai-glm-5-3":{"id":"zai-glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"mistral-medium-2508":{"id":"mistral-medium-2508","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"ministral-3b-latest":{"id":"ministral-3b-latest","name":"Ministral 3B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-01","last_updated":"2024-10-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":128000},"cost":{"input":0.1,"output":0.1,"cache_read":0.01}},"zai-glm-5-2":{"id":"zai-glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"devstral-medium-latest":{"id":"devstral-medium-latest","name":"Devstral 2 (latest)","description":"Legacy model retained for compatibility with older integrations","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"ministral-8b-2512":{"id":"ministral-8b-2512","name":"Ministral 3 8B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-3-8b-instruct-2512"}}},"abliteration-ai":{"id":"abliteration-ai","env":["ABLIT_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.abliteration.ai/v1","name":"abliteration.ai","doc":"https://docs.abliteration.ai/models","models":{"abliterated-model-large":{"id":"abliterated-model-large","name":"Abliterated Model Large","description":"GLM-5.2 model abliterated and finetuned for cyber, ML red teaming, and agent testing","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-25","last_updated":"2026-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":5,"output":5,"cache_read":0.5}},"abliterated-model":{"id":"abliterated-model","name":"Abliterated Model","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-01-06","last_updated":"2026-07-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":150000,"input":150000,"output":8192},"cost":{"input":3,"output":3,"cache_read":0.3}},"abliterated-model-large-v2":{"id":"abliterated-model-large-v2","name":"Abliterated Model Large V2","description":"GLM-5.3 model abliterated and finetuned for cyber, ML red teaming, and agent testing","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-31","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":5,"output":5,"cache_read":0.5}}}},"openai":{"id":"openai","env":["OPENAI_API_KEY"],"npm":"@ai-sdk/openai","name":"OpenAI","doc":"https://platform.openai.com/docs/models","models":{"o1-pro":{"id":"o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":150,"output":600}},"gpt-3.5-turbo":{"id":"gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"status":"deprecated","cost":{"input":0.5,"output":1.5,"cache_read":0}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-daybreak-blue-latest":{"id":"gpt-daybreak-blue-latest","name":"Daybreak Blue","description":"Alias for flagship general-purpose models with safeguards for defensive cybersecurity work","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-5.2-pro":{"id":"gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168}},"gpt-4o-2024-05-13":{"id":"gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":8.75,"output":26.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":15}},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}},"ultrafast":{"cost":{"input":12,"output":60,"cache_read":0.6,"cache_write":15},"provider":{"body":{"service_tier":"ultrafast"}}}}},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"gpt-4o-2024-08-06":{"id":"gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25}},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"gpt-image-1-mini":{"id":"gpt-image-1-mini","type":"image","name":"gpt-image-1-mini","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0}},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.2,"output":1,"cache_read":0.02,"cache_write":0.25},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":0.2,"output":0.8,"cache_read":0.05},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.1,"output":0.4,"cache_read":0.025}},"chatgpt-image-latest":{"id":"chatgpt-image-latest","type":"image","name":"chatgpt-image-latest","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0}},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}}},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"gpt-image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-01","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":1536},"cost":{"input":0.02,"output":0}},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":0.25,"output":1,"cache_read":0.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.6,"cache_read":0.075}},"gpt-5.6":{"id":"gpt-5.6","name":"GPT-5.6","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-4":{"id":"gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"status":"deprecated","cost":{"input":30,"output":60}},"gpt-image-1":{"id":"gpt-image-1","type":"image","name":"gpt-image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0},"status":"deprecated","canonical_model_id":"openai/gpt-image-1"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.875,"output":7,"cache_read":0.0875},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":0.7,"output":2.8,"cache_read":0.175},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.4,"output":1.6,"cache_read":0.1}},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.125,"output":1,"cache_read":0.0125},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":20,"output":100,"cache_read":2,"cache_write":25},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}},"ultrafast":{"cost":{"input":60,"output":300,"cache_read":6,"cache_write":75},"provider":{"body":{"service_tier":"ultrafast"}}}}},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}}},"gpt-4-turbo":{"id":"gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":10,"output":30}},"gpt-5.3-codex-spark":{"id":"gpt-5.3-codex-spark","name":"GPT-5.3 Codex Spark","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-spark","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":100000,"output":32000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}}},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120}},"gpt-realtime-2.1":{"id":"gpt-realtime-2.1","type":"realtime","name":"GPT-Realtime-2.1","description":"Realtime speech-to-speech model with configurable reasoning, tool use, and robust voice-agent behavior","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","audio","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"input":96000,"output":32000},"cost":{"input":4,"output":24,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-2.1"},"o1":{"id":"o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5}},"gpt-daybreak-red-latest":{"id":"gpt-daybreak-red-latest","name":"Daybreak Red","description":"Alias for advanced cybersecurity models for authorized vulnerability research and security testing","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":12.5,"output":75,"cache_read":1.25,"cache_write":15.625},"canonical_model_id":"openai/gpt-5.6-cyber"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.625,"output":5,"cache_read":0.0625},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"flex":{"cost":{"input":0.1,"output":0.625,"cache_read":0.01},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.2,"output":1.25,"cache_read":0.02}},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":4.25,"output":17,"cache_read":2.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":10,"cache_read":1.25}},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":2,"output":8,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.55,"output":2.2,"cache_read":0.1375},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.1,"output":4.4,"cache_read":0.275}},"o3-pro":{"id":"o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80}},"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"gpt-image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0},"canonical_model_id":"openai/gpt-image-1.5"},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-01","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":3072},"cost":{"input":0.13,"output":0}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.625,"output":5,"cache_read":0.0625},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"flex":{"cost":{"input":0.025,"output":0.2,"cache_read":0.0025},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.05,"output":0.4,"cache_read":0.005}},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":1,"output":4,"cache_read":0.25},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5}},"gpt-4o-2024-11-20":{"id":"gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25}},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}}},"text-embedding-ada-002":{"id":"text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2022-12","release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.1,"output":0}},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5}},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.375,"output":2.25,"cache_read":0.0375},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075}}}},"alibaba-token-plan-cn":{"id":"alibaba-token-plan-cn","env":["ALIBABA_TOKEN_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan.cn-beijing.maas.aliyuncs.com/compatible-mode/v1","name":"Alibaba Token Plan (China)","doc":"https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"qwen-image-2.0-pro":{"id":"qwen-image-2.0-pro","type":"image","name":"Qwen Image 2.0 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-plus"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"happyhorse-1.1-t2v":{"id":"happyhorse-1.1-t2v","type":"video","name":"HappyHorse 1.1 Text-to-Video","description":"Video model for prompt-driven text-to-video generation","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.cn-beijing.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"qwen3.8-max-preview":{"id":"qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"wan2.7-image-pro":{"id":"wan2.7-image-pro","type":"image","name":"Wan2.7 Image Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-max"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"wan2.7-image":{"id":"wan2.7-image","type":"image","name":"Wan2.7 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"input":196601,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M2.5"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"happyhorse-1.1-r2v":{"id":"happyhorse-1.1-r2v","type":"video","name":"HappyHorse 1.1 Reference-to-Video","description":"Video model for reference-guided video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.cn-beijing.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"qwen-image-2.0":{"id":"qwen-image-2.0","type":"image","name":"Qwen Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"happyhorse-1.1-i2v":{"id":"happyhorse-1.1-i2v","type":"video","name":"HappyHorse 1.1 Image-to-Video","description":"Video model for image-to-video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.cn-beijing.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-03","last_updated":"2025-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0,"output":0}},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-flash"}}},"volcengine-coding-plan":{"id":"volcengine-coding-plan","env":["ARK_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ark.cn-beijing.volces.com/api/coding/v3","name":"Volcengine Ark Coding Plan","doc":"https://www.volcengine.com/docs/82379/1928261","models":{"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"doubao-seed-2.0-lite":{"id":"doubao-seed-2.0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"doubao-seed-2.1-turbo":{"id":"doubao-seed-2.1-turbo","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"doubao-seed-evolving":{"id":"doubao-seed-evolving","name":"Seed Evolving","description":"Rolling ByteDance Seed model for rapidly updated reasoning, coding, and agent capabilities","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"bytedance-seed/seed-evolving"}}},"siliconflow":{"id":"siliconflow","env":["SILICONFLOW_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.siliconflow.com/v1","name":"SiliconFlow","doc":"https://cloud.siliconflow.com/models","models":{"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen/Qwen3-30B-A3B-Instruct-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.09,"output":0.3},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":131000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen/Qwen3-32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.57},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":3.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen/Qwen3-VL-30B-A3B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.29,"output":1},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.24,"output":1.8,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-VL-8B-Instruct":{"id":"Qwen/Qwen3-VL-8B-Instruct","name":"Qwen/Qwen3-VL-8B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.18,"output":0.68}},"Qwen/Qwen3-14B":{"id":"Qwen/Qwen3-14B","name":"Qwen/Qwen3-14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.07,"output":0.28}},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen/Qwen3-8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.06,"output":0.06},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3-VL-30B-A3B-Thinking":{"id":"Qwen/Qwen3-VL-30B-A3B-Thinking","name":"Qwen/Qwen3-VL-30B-A3B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-11","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.29,"output":1}},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.24,"output":1.8,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen/Qwen3-Coder-480B-A35B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.25,"output":1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen2.5-7B-Instruct":{"id":"Qwen/Qwen2.5-7B-Instruct","name":"Qwen/Qwen2.5-7B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0.05,"output":0.05},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"Qwen/Qwen3-VL-32B-Thinking":{"id":"Qwen/Qwen3-VL-32B-Thinking","name":"Qwen/Qwen3-VL-32B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.2,"output":1.5}},"Qwen/Qwen3-VL-32B-Instruct":{"id":"Qwen/Qwen3-VL-32B-Instruct","name":"Qwen/Qwen3-VL-32B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.2,"output":0.6}},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen/Qwen3.5-9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen/Qwen3-Coder-30B-A3B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.39,"output":2.34},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen2.5-72B-Instruct":{"id":"Qwen/Qwen2.5-72B-Instruct","name":"Qwen/Qwen2.5-72B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0.59,"output":0.59}},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.26,"output":2.08},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMaxAI/MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":197000,"output":131000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"ByteDance-Seed/Seed-OSS-36B-Instruct":{"id":"ByteDance-Seed/Seed-OSS-36B-Instruct","name":"ByteDance-Seed/Seed-OSS-36B-Instruct","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-04","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.21,"output":0.57}},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.75,"output":1,"cache_read":0.25},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-4-12B-it":{"id":"google/gemma-4-12B-it","name":"Gemma 4 12B IT","description":"Compact Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"google/gemma-4-12b-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"openai/gpt-oss-20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":8000},"cost":{"input":0.04,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"openai/gpt-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":8000},"cost":{"input":0.05,"output":0.45},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.50162,"output":3.135,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.13,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"deepseek-ai/DeepSeek-V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.25,"output":1},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"id":"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"deepseek-ai/DeepSeek-V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"deepseek-ai/DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":0.42,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-V3.2-Exp":{"id":"deepseek-ai/DeepSeek-V3.2-Exp","name":"deepseek-ai/DeepSeek-V3.2-Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-10","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":0.41}},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"deepseek-ai/DeepSeek-R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.5,"output":2.18},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"deepseek-ai/DeepSeek-V3.1-Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":1}},"stepfun-ai/Step-3.5-Flash":{"id":"stepfun-ai/Step-3.5-Flash","name":"stepfun-ai/Step-3.5-Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":66000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"moonshotai/Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"moonshotai/Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.77,"output":3.4,"cache_read":0.14},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262000},"cost":{"input":2.7,"output":13.5,"cache_read":0.27},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.85916,"output":3.8,"cache_read":0.17993},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"inclusionAI/Ling-flash-2.0":{"id":"inclusionAI/Ling-flash-2.0","name":"inclusionAI/Ling-flash-2.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.57}},"tencent/Hunyuan-A13B-Instruct":{"id":"tencent/Hunyuan-A13B-Instruct","name":"tencent/Hunyuan-A13B-Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.57}},"tencent/Hy3":{"id":"tencent/Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":262144},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"nex-agi/Nex-N2-Pro":{"id":"nex-agi/Nex-N2-Pro","name":"Nex-N2-Pro","description":"Open agentic MoE model (397B total, 17B active) for coding, tool use, and research workflows","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"status":"deprecated","cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"nex-agi/nex-n2-pro"},"meituan-longcat/LongCat-2.0":{"id":"meituan-longcat/LongCat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":131072},"cost":{"input":0.75,"output":2.95,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"zai-org/GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131000},"cost":{"input":1.19,"output":3.74,"cache_read":0.6,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"zai-org/GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"cost":{"input":0.95,"output":2.55,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5V-Turbo":{"id":"zai-org/GLM-5V-Turbo","name":"zai-org/GLM-5V-Turbo","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5v-turbo"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"zai-org/GLM-4.5-Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.86},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":1.302,"output":4.092,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"infer":{"id":"infer","env":["INFER_API_KEY"],"npm":"@ai-sdk/openai","api":"https://infer.flow7.org/v1","name":"Infer by Flow7","doc":"https://infer.flow7.org/opencode","models":{"infer/gpt-5.6-sol:official":{"id":"infer/gpt-5.6-sol:official","name":"GPT-5.6 Sol (Official API)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":271999,"output":128000},"provider":{"shape":"responses"},"cost":{"input":2.5,"output":12.5,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-sol"},"infer/gpt-6-astra:official":{"id":"infer/gpt-6-astra:official","name":"GPT-6 Astra (Official API)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":271999,"output":128000},"provider":{"shape":"responses"},"cost":{"input":12.5,"output":62.5,"cache_read":1.25,"cache_write":15.625},"canonical_model_id":"openai/gpt-6-astra"}}},"perplexity":{"id":"perplexity","env":["PERPLEXITY_API_KEY"],"npm":"@ai-sdk/perplexity","name":"Perplexity","doc":"https://docs.perplexity.ai","models":{"sonar-pro":{"id":"sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"sonar-reasoning-pro":{"id":"sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"sonar":{"id":"sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"sonar-deep-research":{"id":"sonar-deep-research","name":"Perplexity Sonar Deep Research","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"}}},"amd":{"id":"amd","env":["AMD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://developer.amd.com.cn/radeon/api/v1","name":"AMD","doc":"https://developer.amd.com.cn/radeon/tokenfactory","models":{"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"DeepSeek-V4-Flash-Vision-Exp":{"id":"DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"MiMo-V2.6-Flash":{"id":"MiMo-V2.6-Flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"DeepSeek-V4.1-Flash":{"id":"DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"MiniCPM-V-4.7-35B-A3B":{"id":"MiniCPM-V-4.7-35B-A3B","name":"MiniCPM-V 4.7 35B-A3B","description":"First MoE MiniCPM-V model for single-image, multi-image and video understanding","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"openbmb/minicpm-v-4.7-35b-a3b"},"MiniCPM5-2B":{"id":"MiniCPM5-2B","name":"MiniCPM5-2B","description":"Dense 2B-class open-source model for on-device and resource-constrained use, with native long-context support, tool calling, and agentic tasks","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-09-06","last_updated":"2026-09-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"openbmb/minicpm5-2b"},"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen3.8-Flash-Next":{"id":"Qwen3.8-Flash-Next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"GLM-5.3-Flash":{"id":"GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"MiniCPM-V-4.7-1B":{"id":"MiniCPM-V-4.7-1B","name":"MiniCPM-V 4.7 1B","description":"Compact MiniCPM-V model for single-image, multi-image and video understanding","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"openbmb/minicpm-v-4.7-1b"},"Gemma-4-31B-IT":{"id":"Gemma-4-31B-IT","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"google/gemma-4-31b-it"}}},"arcee":{"id":"arcee","env":["ARCEE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.arcee.ai/api/v1","name":"Arcee","doc":"https://docs.arcee.ai","models":{"trinity-large-thinking":{"id":"trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":80000},"status":"beta","cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"beta","cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":384000},"status":"deprecated","cost":{"input":1.74,"output":3.48,"cache_read":0.2},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"beta","cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"beta","cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-latest":{"id":"deepseek/deepseek-v4-flash-latest","name":"DeepSeek V4 Flash Latest","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"beta","cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"beta","cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"zai-org/glm-5.3":{"id":"zai-org/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"beta","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"status":"beta","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-5.3-flash":{"id":"zai-org/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"beta","cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"zeldoc":{"id":"zeldoc","env":["ZELDOC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.zeldoc.ai/v1","name":"Zeldoc","doc":"https://docs.zeldoc.ai","models":{"zdev":{"id":"zdev","name":"ZDev","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0}}}},"cloudflare-workers-ai":{"id":"cloudflare-workers-ai","env":["CLOUDFLARE_ACCOUNT_ID","CLOUDFLARE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/ai/v1","name":"Cloudflare Workers AI","doc":"https://developers.cloudflare.com/workers-ai/models/","models":{"@cf/ibm-granite/granite-4.0-h-micro":{"id":"@cf/ibm-granite/granite-4.0-h-micro","name":"Granite 4.0 H Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.017,"output":0.112},"canonical_model_id":"ibm/granite-4-h-micro"},"@cf/aisingapore/gemma-sea-lion-v4-27b-it":{"id":"@cf/aisingapore/gemma-sea-lion-v4-27b-it","name":"Gemma Sea Lion V4 27B It","description":"Gemma 3 27B tuned by AI Singapore for Southeast Asian languages and instruction following","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"aisingapore/gemma-sea-lion-v4-27b-it"},"@cf/google/gemma-4-26b-a4b-it":{"id":"@cf/google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"@cf/openai/gpt-oss-20b":{"id":"@cf/openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.2,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"@cf/openai/gpt-oss-120b":{"id":"@cf/openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"@cf/deepseek-ai/deepseek-r1-distill-qwen-32b":{"id":"@cf/deepseek-ai/deepseek-r1-distill-qwen-32b","name":"Deepseek R1 Distill Qwen 32B","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":80000},"cost":{"input":0.497,"output":4.881},"canonical_model_id":"deepseek/deepseek-r1"},"@cf/deepseek-ai/deepseek-v4.1-flash-1017":{"id":"@cf/deepseek-ai/deepseek-v4.1-flash-1017","name":"Deepseek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.006}},"@cf/deepseek-ai/deepseek-v4-pro-0813":{"id":"@cf/deepseek-ai/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"@cf/deepseek-ai/deepseek-v4-flash-0731":{"id":"@cf/deepseek-ai/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"@cf/moonshotai/kimi-k2.7-code":{"id":"@cf/moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"@cf/moonshotai/kimi-k2.6":{"id":"@cf/moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"@cf/meta/llama-3.2-11b-vision-instruct":{"id":"@cf/meta/llama-3.2-11b-vision-instruct","name":"Llama 3.2 11B Vision Instruct","description":"Open Llama multimodal model for image understanding and text reasoning","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.0485,"output":0.676},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"@cf/meta/llama-4-scout-17b-16e-instruct":{"id":"@cf/meta/llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B 16E Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":16384},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"@cf/meta/llama-guard-3-8b":{"id":"@cf/meta/llama-guard-3-8b","name":"Llama Guard 3 8B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.484,"output":0.03},"canonical_model_id":"meta/llama-guard-3-8b"},"@cf/meta/llama-3.2-3b-instruct":{"id":"@cf/meta/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":80000},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"meta/llama-3.2-3b"},"@cf/meta/llama-3.3-70b-instruct-fp8-fast":{"id":"@cf/meta/llama-3.3-70b-instruct-fp8-fast","name":"Llama 3.3 70B Instruct fp8 Fast","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":24000,"output":24000},"cost":{"input":0.293,"output":2.253},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"@cf/meta/llama-3.1-8b-instruct-fp8":{"id":"@cf/meta/llama-3.1-8b-instruct-fp8","name":"Llama 3.1 8B Instruct fp8","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.152,"output":0.287},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"@cf/meta/llama-3.2-1b-instruct":{"id":"@cf/meta/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":60000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b"},"@cf/mistralai/mistral-small-3.1-24b-instruct":{"id":"@cf/mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B Instruct","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"@cf/qwen/qwen3.8-27b":{"id":"@cf/qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":3.2,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"@cf/qwen/qwen2.5-coder-32b-instruct":{"id":"@cf/qwen/qwen2.5-coder-32b-instruct","name":"Qwen2.5 Coder 32B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"@cf/qwen/qwen3-30b-a3b-fp8":{"id":"@cf/qwen/qwen3-30b-a3b-fp8","name":"Qwen3 30B A3b fp8","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"@cf/qwen/qwq-32b":{"id":"@cf/qwen/qwq-32b","name":"Qwq 32B","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":24000,"output":24000},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwq-32b"},"@cf/nvidia/nemotron-3-120b-a12b":{"id":"@cf/nvidia/nemotron-3-120b-a12b","name":"Nemotron 3 Super 120B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"@cf/zai-org/glm-5.3":{"id":"@cf/zai-org/glm-5.3","name":"Glm 5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"@cf/zai-org/glm-4.7-flash":{"id":"@cf/zai-org/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"@cf/zai-org/glm-5.2":{"id":"@cf/zai-org/glm-5.2","name":"Glm 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"@cf/zai-org/glm-5.3-flash":{"id":"@cf/zai-org/glm-5.3-flash","name":"Glm 5.3 Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"stepfun-ai":{"id":"stepfun-ai","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.ai/v1","name":"StepFun (Global)","doc":"https://platform.stepfun.ai/docs/en/overview/concept","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"step-tts-2":{"id":"step-tts-2","type":"speech","name":"Step TTS 2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-01","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"input":1024000,"output":65536},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"stepaudio-2.5-asr":{"id":"stepaudio-2.5-asr","type":"transcription","name":"StepAudio 2.5 ASR","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"step","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-24","last_updated":"2026-07-02","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"stepaudio-2.5-tts":{"id":"stepaudio-2.5-tts","type":"speech","name":"StepAudio 2.5 TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}}}},"regolo-ai":{"id":"regolo-ai","env":["REGOLO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.regolo.ai/v1","name":"Regolo AI","doc":"https://docs.regolo.ai/","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":120000,"output":120000},"cost":{"input":0.58,"output":2.42},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen-image":{"id":"qwen-image","type":"image","name":"Qwen-Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-01","last_updated":"2026-03-01","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":4096},"cost":{"input":0.5,"output":2}},"deepseek-ocr-2":{"id":"deepseek-ocr-2","name":"DeepSeek OCR 2","description":"High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":4000,"output":4000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-ocr-2"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.6,"output":2.7},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"brick-complexity-pro":{"id":"brick-complexity-pro","name":"Brick Complexity Pro","description":"Complexity classifier that powers the Brick semantic router by extracting query difficulty","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-06","last_updated":"2026-02-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":100000,"output":15000},"cost":{"input":0.12,"output":0.46}},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT-OSS-20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.11215,"output":0.47103},"canonical_model_id":"openai/gpt-oss-20b"},"apertus-70b":{"id":"apertus-70b","name":"Apertus 70B","description":"Fully open 70B multilingual LLM supporting 1800+ languages with 65K context. Trained on 15T tokens of compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":30000,"output":30000},"cost":{"input":0.46,"output":2.42},"canonical_model_id":"swiss-ai/apertus-70b"},"qwen3-embedding-8b":{"id":"qwen3-embedding-8b","type":"embedding","name":"Qwen3-Embedding-8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5-9B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.078505,"output":0.392525},"canonical_model_id":"alibaba/qwen3.5-9b"},"glm5.2":{"id":"glm5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":96000,"output":96000},"cost":{"input":2.31,"output":6},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.5-122b":{"id":"qwen3.5-122b","name":"Qwen3.5-122B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":1.1215,"output":4.7103},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT-OSS-120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":1.1215,"output":4.7103},"canonical_model_id":"openai/gpt-oss-120b"},"gemma4-31b":{"id":"gemma4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":100000},"cost":{"input":0.46,"output":2.42},"canonical_model_id":"google/gemma-4-31b-it"},"faster-whisper-large-v3":{"id":"faster-whisper-large-v3","type":"transcription","name":"Faster Whisper Large v3","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"openai/whisper-large-v3"},"qwen3-reranker-4b":{"id":"qwen3-reranker-4b","type":"reranking","name":"Qwen3-Reranker-4B","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-01","last_updated":"2026-02-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.12,"output":0.12}},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3-Coder-Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.3,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"brick-v1-beta":{"id":"brick-v1-beta","name":"Brick v1 Beta","description":"Semantic router by Regolo.ai that directs each request to the most suitable model, optimizing costs and performance","family":"model-router","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-06","last_updated":"2026-02-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":100000,"output":15000},"status":"beta","cost":{"input":0,"output":0}},"mistral-small-4-119b":{"id":"mistral-small-4-119b","name":"Mistral Small 4 119B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.56075,"output":2.35515},"canonical_model_id":"mistral/mistral-small-2603"}}},"huggingface":{"id":"huggingface","env":["HF_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://router.huggingface.co/v1","name":"Hugging Face","doc":"https://huggingface.co/docs/inference-providers","models":{"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":6.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.29,"output":0.59},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.47,"output":3.19},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3-VL-235B-A22B-Thinking":{"id":"Qwen/Qwen3-VL-235B-A22B-Thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.855,"output":2.565},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3-235B-A22B-Thinking-2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.3,"output":3},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3-30B-A3B":{"id":"Qwen/Qwen3-30B-A3B","name":"Qwen3 30B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"Qwen/Qwen3-VL-235B-A22B-Instruct":{"id":"Qwen/Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen3 8B","description":"Dense 8B Qwen3 model with switchable thinking for chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":32768},"cost":{"input":0.07,"output":0.18},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3-Coder-Next":{"id":"Qwen/Qwen3-Coder-Next","name":"Qwen3-Coder-Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.15,"output":0.95},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen3-Coder-480B-A35B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":2,"output":2},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3-235B-A22B":{"id":"Qwen/Qwen3-235B-A22B","name":"Qwen3 235B-A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"Qwen/Qwen2.5-Coder-32B-Instruct":{"id":"Qwen/Qwen2.5-Coder-32B-Instruct","name":"Qwen2.5-Coder-32B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.06,"output":0.2},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next-80B-A3B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.25,"output":1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":3},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3-Embedding-4B":{"id":"Qwen/Qwen3-Embedding-4B","type":"embedding","name":"Qwen 3 Embedding 4B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-12","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://router.huggingface.co/deepinfra/v1/openai"},"cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-4b"},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.07,"output":0.26},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"Qwen/Qwen3-Embedding-8B":{"id":"Qwen/Qwen3-Embedding-8B","type":"embedding","name":"Qwen 3 Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-12","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://router.huggingface.co/deepinfra/v1/openai"},"cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5-397B-A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M2":{"id":"MiniMaxAI/MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"MiniMaxAI/MiniMax-M1-80k":{"id":"MiniMaxAI/MiniMax-M1-80k","name":"MiniMax M1 80K","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":80000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"minimax/minimax-m1-80k"},"MiniMaxAI/MiniMax-M2.1":{"id":"MiniMaxAI/MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-10","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":4.05},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"thinkingmachines/inkling-small"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.25,"output":0.69},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":8192},"cost":{"input":0.4,"output":1.3},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"id":"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.27,"output":1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek-R1-0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":3,"output":5},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.28,"output":0.4},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":32768},"cost":{"input":0.7,"output":2.5},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.27,"output":1.12},"canonical_model_id":"deepseek/deepseek-v3-0324"},"stepfun-ai/Step-3.5-Flash":{"id":"stepfun-ai/Step-3.5-Flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"stepfun-ai/Step-3.7-Flash":{"id":"stepfun-ai/Step-3.7-Flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15},"canonical_model_id":"stepfun/step-3.7-flash"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2-Thinking":{"id":"moonshotai/Kimi-K2-Thinking","name":"Kimi-K2-Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/Kimi-K2-Instruct-0905":{"id":"moonshotai/Kimi-K2-Instruct-0905","name":"Kimi-K2-Instruct-0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":1,"output":3},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/Kimi-K2-Instruct":{"id":"moonshotai/Kimi-K2-Instruct","name":"Kimi-K2-Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1,"output":3},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"tencent/Hy4-preview":{"id":"tencent/Hy4-preview","name":"Hy4 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501},"canonical_model_id":"tencent/hy4-preview"},"tencent/Hy3":{"id":"tencent/Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.14,"output":0.58},"canonical_model_id":"tencent/hy3"},"meta-llama/Llama-3.1-8B-Instruct":{"id":"meta-llama/Llama-3.1-8B-Instruct","name":"Llama-3.1-8B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.06,"output":0.06},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/Llama-Guard-4-12B":{"id":"meta-llama/Llama-Guard-4-12B","name":"Llama Guard 4 12B","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"XiaomiMiMo/MiMo-V2.5":{"id":"XiaomiMiMo/MiMo-V2.5","name":"MiMo-V2.5","description":"MiMo model for long-context reasoning, perception, and agentic tasks","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.4,"output":2},"canonical_model_id":"xiaomi/mimo-v2.5"},"XiaomiMiMo/MiMo-V2.5-Pro":{"id":"XiaomiMiMo/MiMo-V2.5-Pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":3},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-4.6V-Flash":{"id":"zai-org/GLM-4.6V-Flash","name":"GLM-4.6V-Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-4.7-Flash":{"id":"zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/AutoGLM-Phone-9B-Multilingual":{"id":"zai-org/AutoGLM-Phone-9B-Multilingual","name":"AutoGLM Phone 9B Multilingual","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-12-10","last_updated":"2025-12-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.035,"output":0.138},"canonical_model_id":"zhipuai/autoglm-phone-9b-multilingual"},"zai-org/GLM-4.5V":{"id":"zai-org/GLM-4.5V","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8},"canonical_model_id":"zhipuai/glm-4.5v"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":1.4,"output":4.4},"canonical_model_id":"zhipuai/glm-5.2"}}},"ai21":{"id":"ai21","env":["AI21_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ai21.com/studio/v1","name":"AI21 Labs","doc":"https://docs.ai21.com/docs/jamba-foundation-models","models":{"jamba-mini":{"id":"jamba-mini","name":"Jamba Mini","description":"AI21's efficient, lightweight hybrid SSM-Transformer model for a wide range of tasks","family":"jamba","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-22","release_date":"2026-01-08","last_updated":"2026-01-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"ai21/jamba-mini"},"jamba-large":{"id":"jamba-large","name":"Jamba Large","description":"AI21's hybrid SSM-Transformer long-context model for enterprise agents and grounded generation","family":"jamba","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-22","release_date":"2025-07-03","last_updated":"2025-07-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"ai21/jamba-large"}}},"pendra":{"id":"pendra","env":["PENDRA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.pendra.ai/api/v1","name":"Pendra","doc":"https://pendra.ai/docs/integrations/opencode","models":{"qwen3-coder:30b":{"id":"qwen3-coder:30b","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"llama3.3:70b":{"id":"llama3.3:70b","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"qwen3.6:27b":{"id":"qwen3.6:27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"gpt-oss:120b":{"id":"gpt-oss:120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-120b"}}},"blueclaw":{"id":"blueclaw","env":["BLUECLAW_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://openai.blueclaw.network/v1","name":"Blue Claw","doc":"https://blueclaw.network","models":{"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"beta","canonical_model_id":"alibaba/qwen3.8-27b"}}},"moonshotai":{"id":"moonshotai","env":["MOONSHOT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.moonshot.ai/v1","name":"Moonshot AI","doc":"https://platform.moonshot.ai/docs/api/chat","models":{"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"kimi-k3.1-pro-highspeed":{"id":"kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code HighSpeed","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code"}}},"inference":{"id":"inference","env":["INFERENCE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.net/v1","name":"Inference","doc":"https://inference.net/models","models":{"google/gemma-3":{"id":"google/gemma-3","name":"Google Gemma 3","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":125000,"output":4096},"cost":{"input":0.15,"output":0.3}},"mistral/mistral-nemo-12b-instruct":{"id":"mistral/mistral-nemo-12b-instruct","name":"Mistral Nemo 12B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.038,"output":0.1},"canonical_model_id":"mistral/mistral-nemo"},"meta/llama-3.2-11b-vision-instruct":{"id":"meta/llama-3.2-11b-vision-instruct","name":"Llama 3.2 11B Vision Instruct","description":"Open Llama multimodal model for image understanding and text reasoning","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.055,"output":0.055},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"meta/llama-3.1-8b-instruct":{"id":"meta/llama-3.1-8b-instruct","name":"Llama 3.1 8B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.025,"output":0.025},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta/llama-3.2-3b-instruct":{"id":"meta/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"cost":{"input":0.02,"output":0.02}},"meta/llama-3.2-1b-instruct":{"id":"meta/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"qwen/qwen-2.5-7b-vision-instruct":{"id":"qwen/qwen-2.5-7b-vision-instruct","name":"Qwen 2.5 7B Vision Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":125000,"output":4096},"status":"deprecated","cost":{"input":0.2,"output":0.2}},"qwen/qwen3-embedding-4b":{"id":"qwen/qwen3-embedding-4b","type":"embedding","name":"Qwen 3 Embedding 4B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-12","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":2048},"status":"deprecated","cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-4b"},"osmosis/osmosis-structure-0.6b":{"id":"osmosis/osmosis-structure-0.6b","name":"Osmosis Structure 0.6B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"osmosis","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4000,"output":2048},"status":"deprecated","cost":{"input":0.1,"output":0.5}}}},"meta":{"id":"meta","env":["META_MODEL_API_KEY"],"npm":"@ai-sdk/openai","api":"https://api.meta.ai/v1","name":"Meta","doc":"https://dev.meta.ai/docs","models":{"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"muse-spark-1.2-contributor":{"id":"muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"}}},"qvac":{"id":"qvac","env":["QVAC_API_KEY"],"npm":"@qvac/ai-sdk-provider","name":"QVAC","doc":"https://www.npmjs.com/package/@qvac/ai-sdk-provider","models":{"qwen3.5-4b":{"id":"qwen3.5-4b","name":"Qwen3.5 4B","description":"Qwen instruction model for multilingual chat and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-4b"},"qwen3.5-2b":{"id":"qwen3.5-2b","name":"Qwen3.5 2B","description":"Qwen instruction model for multilingual chat and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-2b"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-20b"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-120b"},"gemma4-31b":{"id":"gemma4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"qwen3.5-0.8b":{"id":"qwen3.5-0.8b","name":"Qwen3.5 0.8B","description":"Qwen instruction model for multilingual chat and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-0.8b"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"tinfoil":{"id":"tinfoil","env":["TINFOIL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.tinfoil.sh/v1","name":"Tinfoil","doc":"https://docs.tinfoil.sh","models":{"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.8},"canonical_model_id":"moonshotai/kimi-k3"},"gpt-oss-safeguard-120b":{"id":"gpt-oss-safeguard-120b","name":"gpt-oss-safeguard-120b","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"beta","cost":{"input":0.65,"output":1.45,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"nomic-embed-text":{"id":"nomic-embed-text","name":"Nomic Embed Text v1.5","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2024-02","last_updated":"2024-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":768},"cost":{"input":0.05,"output":0}},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"llama3-3-70b":{"id":"llama3-3-70b","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":1.75,"output":2.75},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gemma4-31b":{"id":"gemma4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":1},"canonical_model_id":"google/gemma-4-31b-it"},"glm-5-3":{"id":"glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.8,"output":5.75,"cache_read":0.45},"canonical_model_id":"zhipuai/glm-5.3"}}},"openreason":{"id":"openreason","env":["OPENREASON_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.openreason.app/v1","name":"OpenReason","doc":"https://openreason.app/docs","models":{"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1055,"output":0.422},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/deepseek-v4-flash-0731":{"id":"deepseek-ai/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.1371,"output":0.2743},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.0022,"output":4.22},"canonical_model_id":"moonshotai/kimi-k2.7-code"}}},"ai-router":{"id":"ai-router","env":["AI_ROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ai-router.dev/v1","name":"AI-ROUTER","doc":"https://ai-router.dev/openai-compatible-api-gateway/","models":{"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":1,"output":6,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"openai/gpt-5.6-sol"}}},"io-net":{"id":"io-net","env":["IOINTELLIGENCE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.intelligence.io.solutions/api/v1","name":"IO.NET","doc":"https://io.net/docs/guides/intelligence/io-intelligence","models":{"Qwen/Qwen2.5-VL-32B-Instruct":{"id":"Qwen/Qwen2.5-VL-32B-Instruct","name":"Qwen 2.5 VL 32B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-11-01","last_updated":"2024-11-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"status":"deprecated","cost":{"input":0.05,"output":0.22,"cache_read":0.025,"cache_write":0.1}},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen 3 235B Thinking","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"status":"deprecated","cost":{"input":0.11,"output":0.6,"cache_read":0.055,"cache_write":0.22},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen 3 Next 80B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"cost":{"input":0.11875,"output":1.2625,"cache_read":0.059375},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"Intel/Qwen3-Coder-480B-A35B-Instruct-int4-mixed-ar":{"id":"Intel/Qwen3-Coder-480B-A35B-Instruct-int4-mixed-ar","name":"Qwen 3 Coder 480B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":106000,"output":4096},"cost":{"input":0.46125,"output":2.29375,"cache_read":0.230625},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT-OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":4096},"cost":{"input":0.055,"output":0.194,"cache_read":0.0275},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT-OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.188,"output":0.7,"cache_read":0.094},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.523667,"output":2.205333,"cache_read":0.261834},"canonical_model_id":"deepseek/deepseek-r1-0528"},"moonshotai/Kimi-K2-Thinking":{"id":"moonshotai/Kimi-K2-Thinking","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":0.6,"output":2.5,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/Kimi-K2-Instruct-0905":{"id":"moonshotai/Kimi-K2-Instruct-0905","name":"Kimi K2 Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":0.57,"output":2.3,"cache_read":0.285},"canonical_model_id":"moonshotai/kimi-k2-0905"},"mistralai/Mistral-Nemo-Instruct-2407":{"id":"mistralai/Mistral-Nemo-Instruct-2407","name":"Mistral Nemo Instruct 2407","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-05","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.0514,"output":0.082,"cache_read":0.0257},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/Magistral-Small-2506":{"id":"mistralai/Magistral-Small-2506","name":"Magistral Small 2506","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.5,"output":1.5,"cache_read":0.25,"cache_write":1},"canonical_model_id":"mistral/magistral-small-2506"},"mistralai/Mistral-Large-Instruct-2411":{"id":"mistralai/Mistral-Large-Instruct-2411","name":"Mistral Large Instruct 2411","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-11-18","last_updated":"2024-11-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":2,"output":6,"cache_read":1,"cache_write":4},"canonical_model_id":"mistral/mistral-large-2411"},"mistralai/Devstral-Small-2505":{"id":"mistralai/Devstral-Small-2505","name":"Devstral Small 2505","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-05-01","last_updated":"2025-05-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.05,"output":0.22,"cache_read":0.025,"cache_write":0.1}},"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"id":"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","name":"Llama 4 Maverick 17B 128E Instruct","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":430000,"output":4096},"cost":{"input":0.26875,"output":0.82625,"cache_read":0.134375},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/Llama-3.2-90B-Vision-Instruct":{"id":"meta-llama/Llama-3.2-90B-Vision-Instruct","name":"Llama 3.2 90B Vision Instruct","description":"Open Llama multimodal model for image understanding and text reasoning","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.35,"output":0.4,"cache_read":0.175,"cache_write":0.7}},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.5466,"output":1.0226,"cache_read":0.2733},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM 4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":4096},"cost":{"input":0.52,"output":2.0375,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-4.6"}}},"scaleway":{"id":"scaleway","env":["SCALEWAY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.scaleway.ai/v1","name":"Scaleway","doc":"https://www.scaleway.com/en/docs/generative-apis/","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"beta","cost":{"input":0.684,"output":3.762,"reasoning":3.762,"cache_read":0.137},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"status":"beta","cost":{"input":0.672,"output":2.016,"reasoning":2.016,"cache_read":0.134},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"status":"beta","cost":{"input":2.01,"output":6.15},"canonical_model_id":"zhipuai/glm-5.3"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":16384},"cost":{"input":1.00935,"output":1.00935},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"whisper-large-v3":{"id":"whisper-large-v3","type":"transcription","name":"Whisper Large v3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2023-09","release_date":"2023-09-01","last_updated":"2026-03-17","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":8192}},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":260000,"output":16384},"cost":{"input":0.841125,"output":2.523375},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen3-embedding-8b":{"id":"qwen3-embedding-8b","type":"embedding","name":"Qwen3 Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0.11215,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"status":"beta","cost":{"input":0.280375,"output":0.56075},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"Mistral Small 3.2 24B Instruct (2506)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.168225,"output":0.392525},"canonical_model_id":"mistral/mistral-small-2506"},"mistral-medium-3.5-128b":{"id":"mistral-medium-3.5-128b","name":"Mistral Medium 3.5 128B","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":1.68225,"output":8.41125},"canonical_model_id":"mistral/mistral-medium-2604"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.6729,"output":4.0374},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"status":"beta","cost":{"input":0.468,"output":0.936,"reasoning":0.936,"cache_read":0.0936},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"lightonocr-3-0.8b":{"id":"lightonocr-3-0.8b","name":"LightOnOCR-3 0.8B","description":"LightOnOCR-3-0.8B is the lightest model of LightOn's LightOnOCR-3 family, with the same layout and visual understanding features as its larger siblings.","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-08","last_updated":"2026-10-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":2.464,"output":0},"canonical_model_id":"lightonai/lightonocr-3-0.8b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":2.0187,"output":6.16825},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT-OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.168225,"output":0.6729},"canonical_model_id":"openai/gpt-oss-120b"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"beta","cost":{"input":0.280375,"output":1.68225},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"bge-multilingual-gemma2":{"id":"bge-multilingual-gemma2","type":"embedding","name":"BGE Multilingual Gemma2","description":"Multilingual Gemma 2 based embedding model for retrieval and semantic search","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-07-25","last_updated":"2024-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8191,"output":0},"cost":{"input":0.11215,"output":0},"canonical_model_id":"baai/bge-multilingual-gemma2"}}},"tencent-token-plan":{"id":"tencent-token-plan","env":["TENCENT_TOKEN_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lkeap.cloud.tencent.com/plan/v3","name":"Tencent Token Plan","doc":"https://cloud.tencent.com/document/product/1823/130060","models":{"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"tencent/hy3"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"}}},"stackit":{"id":"stackit","env":["STACKIT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.openai-compat.model-serving.eu01.onstackit.cloud/v1","name":"STACKIT","doc":"https://docs.stackit.cloud/products/data-and-ai/ai-model-serving/basics/available-shared-models","models":{"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.53,"output":0.76},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3-VL-235B-A22B-Instruct-FP8":{"id":"Qwen/Qwen3-VL-235B-A22B-Instruct-FP8","name":"Qwen3-VL 235B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":218000,"output":16384},"cost":{"input":1.76,"output":2.05},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"Qwen/Qwen3-VL-Embedding-8B":{"id":"Qwen/Qwen3-VL-Embedding-8B","type":"embedding","name":"Qwen3-VL Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.09,"output":0}},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":37000,"output":4096},"status":"deprecated","cost":{"input":0.53,"output":0.76},"canonical_model_id":"google/gemma-3-27b-it"},"intfloat/e5-mistral-7b-instruct":{"id":"intfloat/e5-mistral-7b-instruct","type":"embedding","name":"E5 Mistral 7B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.02243,"output":0}},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.18,"output":0.29},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":8192},"cost":{"input":0.53,"output":0.76},"canonical_model_id":"openai/gpt-oss-120b"},"cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic":{"id":"cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.53,"output":0.76},"canonical_model_id":"meta/llama-3.3-70b-instruct"}}},"poe":{"id":"poe","env":["POE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.poe.com/v1","name":"Poe","doc":"https://creator.poe.com/docs/external-applications/openai-compatible-api","models":{"xai/grok-4.1-fast-non-reasoning":{"id":"xai/grok-4.1-fast-non-reasoning","name":"Grok-4.1-Fast-Non-Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast"},"xai/grok-3-mini":{"id":"xai/grok-3-mini","name":"Grok 3 Mini","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"release_date":"2025-04-11","last_updated":"2025-04-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.3,"output":0.5,"cache_read":0.075}},"xai/grok-4.20-multi-agent":{"id":"xai/grok-4.20-multi-agent","name":"Grok-4.20-Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"temperature":false,"release_date":"2026-03-13","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":0},"cost":{"input":1.2626,"output":2.5253,"cache_read":0.202}},"xai/grok-code-fast-1":{"id":"xai/grok-code-fast-1","name":"Grok Code Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.2,"output":1.5,"cache_read":0.02},"canonical_model_id":"xai/grok-code-fast-1"},"xai/grok-3":{"id":"xai/grok-3","name":"Grok 3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-04-11","last_updated":"2025-04-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.75}},"xai/grok-4-fast-reasoning":{"id":"xai/grok-4-fast-reasoning","name":"Grok-4-Fast-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"status":"deprecated","cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-reasoning"},"xai/grok-4.1-fast-reasoning":{"id":"xai/grok-4.1-fast-reasoning","name":"Grok-4.1-Fast-Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast-reasoning"},"xai/grok-4":{"id":"xai/grok-4","name":"Grok-4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-07-10","last_updated":"2025-07-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.75}},"xai/grok-4-fast-non-reasoning":{"id":"xai/grok-4-fast-non-reasoning","name":"Grok-4-Fast-Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"status":"deprecated","cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"topazlabs-co/topazlabs":{"id":"topazlabs-co/topazlabs","type":"image","name":"TopazLabs","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"topazlabs","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":204,"output":0}},"cerebras/llama-3.1-8b-cs":{"id":"cerebras/llama-3.1-8b-cs","name":"Llama-3.1-8B-CS","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":0},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"cerebras/qwen3-32b-cs":{"id":"cerebras/qwen3-32b-cs","name":"qwen3-32b-cs","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"status":"deprecated","canonical_model_id":"alibaba/qwen3-32b"},"cerebras/qwen3-235b-2507-cs":{"id":"cerebras/qwen3-235b-2507-cs","name":"qwen3-235b-2507-cs","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"status":"deprecated","canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"cerebras/gpt-oss-120b-cs":{"id":"cerebras/gpt-oss-120b-cs","name":"GPT-OSS-120B-CS","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":0},"cost":{"input":0.3535,"output":0.7576},"canonical_model_id":"openai/gpt-oss-120b"},"cerebras/llama-3.3-70b-cs":{"id":"cerebras/llama-3.3-70b-cs","name":"llama-3.3-70b-cs","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"status":"deprecated","canonical_model_id":"meta/llama-3.3-70b-instruct"},"stabilityai/stablediffusionxl":{"id":"stabilityai/stablediffusionxl","type":"image","name":"StableDiffusionXL","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"stable-diffusion","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2023-07-09","last_updated":"2023-07-09","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":200,"output":0}},"runwayml/runway-gen-4-turbo":{"id":"runwayml/runway-gen-4-turbo","type":"video","name":"Runway-Gen-4-Turbo","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"runway","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-09","last_updated":"2025-05-09","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":256,"output":0},"status":"deprecated"},"runwayml/runway":{"id":"runwayml/runway","type":"video","name":"Runway","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"runway","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-10-11","last_updated":"2024-10-11","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":256,"output":0},"status":"deprecated"},"elevenlabs/elevenlabs-v2.5-turbo":{"id":"elevenlabs/elevenlabs-v2.5-turbo","type":"speech","name":"ElevenLabs-v2.5-Turbo","description":"Low-latency multilingual text-to-speech model, superseded by Flash v2.5","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-07-19","last_updated":"2024-07-19","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"elevenlabs/eleven_turbo_v2_5"},"elevenlabs/elevenlabs-music":{"id":"elevenlabs/elevenlabs-music","name":"ElevenLabs-Music","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-08-29","last_updated":"2025-08-29","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":2000,"output":0}},"elevenlabs/elevenlabs-v3":{"id":"elevenlabs/elevenlabs-v3","type":"speech","name":"ElevenLabs-v3","description":"Expressive text-to-speech model with inline audio tags across 70+ languages","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-08-20","last_updated":"2025-08-20","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"elevenlabs/eleven_v3"},"poetools/claude-code":{"id":"poetools/claude-code","name":"claude-code","description":"Claude model for careful reasoning, writing, coding, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"release_date":"2025-11-27","last_updated":"2025-11-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"google/lyria":{"id":"google/lyria","name":"Lyria","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-06-04","last_updated":"2025-06-04","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0}},"google/imagen-3":{"id":"google/imagen-3","type":"image","name":"Imagen-3","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-10-15","last_updated":"2024-10-15","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini-3.1-Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.2525,"output":1.5152},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/veo-3":{"id":"google/veo-3","type":"video","name":"Veo-3","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-21","last_updated":"2025-05-21","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini-2.5-Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":32768}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1065535,"output":65535},"status":"deprecated","cost":{"input":0.87,"output":7,"cache_read":0.087},"canonical_model_id":"google/gemini-2.5-pro"},"google/veo-3-fast":{"id":"google/veo-3-fast","type":"video","name":"Veo-3-Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-13","last_updated":"2025-10-13","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini-3.6-Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.7576,"output":3.7879,"cache_read":0.0758},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-flash":{"id":"google/gemini-3-flash","name":"Gemini-3-Flash","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.404,"output":2.4242,"cache_read":0.0404},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini-3.5-Flash-Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.303,"output":2.5253,"cache_read":0.0303},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-omni-1.1-flash":{"id":"google/gemini-omni-1.1-flash","type":"video","name":"Gemini-Omni-1.1-Flash","description":"Video generation and editing model for scene extension, frame interpolation, and reference-guided multimodal control","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":false,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":131072,"output":57920},"cost":{"input":1.5152,"output":9.0909},"canonical_model_id":"google/gemini-omni-1.1-flash"},"google/veo-3.1-fast":{"id":"google/veo-3.1-fast","type":"video","name":"Veo-3.1-Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/veo-3.1":{"id":"google/veo-3.1","type":"video","name":"Veo-3.1","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/imagen-4-ultra":{"id":"google/imagen-4-ultra","type":"image","name":"Imagen-4-Ultra","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-24","last_updated":"2025-05-24","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemma-4-31b":{"id":"google/gemma-4-31b","name":"Gemma-4-31B","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"google/veo-2":{"id":"google/veo-2","type":"video","name":"Veo-2","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-12-02","last_updated":"2024-12-02","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini-3.7-Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.7576,"output":3.7879,"cache_read":0.0758},"canonical_model_id":"google/gemini-3.7-flash"},"google/imagen-4":{"id":"google/imagen-4","type":"image","name":"Imagen-4","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/nano-banana":{"id":"google/nano-banana","type":"image","name":"Nano-Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":0},"status":"deprecated","cost":{"input":0.21,"output":1.8,"cache_read":0.021},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/nano-banana-pro":{"id":"google/nano-banana-pro","type":"image","name":"Nano-Banana-Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":65536,"output":0},"cost":{"input":2.0202,"output":12.1212,"cache_read":0.202},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini-2.5-Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1065535,"output":65535},"status":"deprecated","cost":{"input":0.21,"output":1.8,"cache_read":0.021},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.0-flash-lite":{"id":"google/gemini-2.0-flash-lite","name":"Gemini-2.0-Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":990000,"output":8192},"status":"deprecated","cost":{"input":0.052,"output":0.21},"canonical_model_id":"google/gemini-2.0-flash-lite"},"google/imagen-4-fast":{"id":"google/imagen-4-fast","type":"image","name":"Imagen-4-Fast","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-06-25","last_updated":"2025-06-25","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini-3.8-Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.7576,"output":3.7879,"cache_read":0.0758},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro":{"id":"google/gemini-3.1-pro","name":"Gemini-3.1-Pro","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2.0202,"output":12.1212,"cache_read":0.202},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini-3.5-Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5152,"output":9.0909,"cache_read":0.1515},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-deep-research":{"id":"google/gemini-deep-research","name":"gemini-deep-research","description":"Legacy model retained for compatibility with older integrations","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":0},"status":"deprecated","cost":{"input":1.6,"output":9.6}},"google/gemini-2.0-flash":{"id":"google/gemini-2.0-flash","name":"Gemini-2.0-Flash","description":"Earlier Gemini Flash workhorse for responsive multimodal apps and tool use","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":990000,"output":8192},"status":"deprecated","cost":{"input":0.1,"output":0.42},"canonical_model_id":"google/gemini-2.0-flash"},"google/imagen-3-fast":{"id":"google/imagen-3-fast","type":"image","name":"Imagen-3-Fast","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-10-17","last_updated":"2024-10-17","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini-2.5-Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1024000,"output":64000},"status":"deprecated","cost":{"input":0.07,"output":0.28},"canonical_model_id":"google/gemini-2.5-flash-lite"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":140,"output":540},"canonical_model_id":"openai/o1-pro"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":2048},"status":"deprecated","cost":{"input":0.45,"output":1.4},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1-Codex-Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.22,"output":1.8,"cache_read":0.022},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/sora-2-pro":{"id":"openai/sora-2-pro","type":"video","name":"Sora-2-Pro","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"sora","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"status":"deprecated"},"openai/gpt-4o-mini-search":{"id":"openai/gpt-4o-mini-search","name":"GPT-4o-mini-Search","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-03-11","last_updated":"2025-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.14,"output":0.54}},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2-Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":19.0909,"output":152.7273},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2-Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.6,"output":13,"cache_read":0.16},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/o3-deep-research":{"id":"openai/o3-deep-research","name":"o3-deep-research","description":"Research model for long-horizon investigation, synthesis, and analytical reports","family":"o","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":9,"output":36,"cache_read":2.2},"canonical_model_id":"openai/o3-deep-research"},"openai/gpt-5.1-instant":{"id":"openai/gpt-5.1-instant","name":"GPT-5.1-Instant","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"openai/gpt-5.3-instant":{"id":"openai/gpt-5.3-instant","name":"GPT-5.3-Instant","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"status":"deprecated","cost":{"input":1.6,"output":13,"cache_read":0.16},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"openai/gpt-4o-search":{"id":"openai/gpt-4o-search","name":"GPT-4o-Search","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-03-11","last_updated":"2025-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2.2727,"output":9.0909}},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":0.99,"output":4},"canonical_model_id":"openai/o3-mini"},"openai/gpt-image-1-mini":{"id":"openai/gpt-image-1-mini","type":"image","name":"GPT-Image-1-Mini","description":"Cost-efficient GPT Image 1 variant for image generation and editing","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"canonical_model_id":"openai/gpt-image-1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1-nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.09,"output":0.36,"cache_read":0.022},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2727,"output":13.6364,"cache_read":0.2273},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-image-2":{"id":"openai/gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5.0505,"output":30.303,"cache_read":1.2626},"canonical_model_id":"openai/gpt-image-2"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.5909,"output":12.7273,"cache_read":0.1591},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5-chat":{"id":"openai/gpt-5-chat","name":"GPT-5-Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":16384},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5-chat-latest"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o-mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":124096,"output":4096},"cost":{"input":0.1364,"output":0.5455,"cache_read":0.0682},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1-Codex-Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-image-1":{"id":"openai/gpt-image-1","type":"image","name":"GPT-Image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":128000,"output":0},"status":"deprecated","canonical_model_id":"openai/gpt-image-1"},"openai/gpt-3.5-turbo-instruct":{"id":"openai/gpt-3.5-turbo-instruct","name":"GPT-3.5-Turbo-Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2023-09-20","last_updated":"2023-09-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":3500,"output":1024},"status":"deprecated","cost":{"input":1.4,"output":1.8}},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.5909,"output":12.7273,"cache_read":0.1591},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1-mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.3636,"output":1.4545,"cache_read":0.0909},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2273,"output":1.8182,"cache_read":0.0227},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4-Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":27.2727,"output":163.6364},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4-Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":9,"output":27},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.3-codex-spark":{"id":"openai/gpt-5.3-codex-spark","name":"GPT-5.3-Codex-Spark","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-spark","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":100000,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.3-codex-spark"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5-Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":922000,"output":128000},"cost":{"input":27.2727,"output":163.6364},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-3.5-turbo-raw":{"id":"openai/gpt-3.5-turbo-raw","name":"GPT-3.5-Turbo-Raw","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":4524,"output":2048},"status":"deprecated","cost":{"input":0.45,"output":1.4},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/sora-2":{"id":"openai/sora-2","type":"video","name":"Sora-2","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"sora","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"status":"deprecated"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5-Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":13.6364,"output":109.0909},"canonical_model_id":"openai/gpt-5-pro"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":14,"output":54},"canonical_model_id":"openai/o1"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1364,"output":9.0909,"cache_read":0.1136},"canonical_model_id":"openai/gpt-5.1"},"openai/o4-mini-deep-research":{"id":"openai/o4-mini-deep-research","name":"o4-mini-deep-research","description":"Research model for long-horizon investigation, synthesis, and analytical reports","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.8,"output":7.2,"cache_read":0.45},"canonical_model_id":"openai/o4-mini-deep-research"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4-Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.1818,"output":1.1364,"cache_read":0.0182},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.2-instant":{"id":"openai/gpt-5.2-instant","name":"GPT-5.2-Instant","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":1.6,"output":13,"cache_read":0.16},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"canonical_model_id":"openai/gpt-4o"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":0.99,"output":4,"cache_read":0.25},"canonical_model_id":"openai/o4-mini"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":18.1818,"output":72.7273},"canonical_model_id":"openai/o3-pro"},"openai/gpt-image-1.5":{"id":"openai/gpt-image-1.5","type":"image","name":"gpt-image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"openai/gpt-image-1.5"},"openai/gpt-4o-aug":{"id":"openai/gpt-4o-aug","name":"GPT-4o-Aug","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2.2727,"output":9.0909,"cache_read":1.1364},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1364,"output":9.0909,"cache_read":0.1136},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5-nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.0455,"output":0.3636,"cache_read":0.0045},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1-Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/dall-e-3":{"id":"openai/dall-e-3","type":"image","name":"DALL-E-3","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"dall-e","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":800,"output":0},"status":"deprecated"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.8182,"output":7.2727,"cache_read":0.4545},"canonical_model_id":"openai/o3"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":922000,"output":128000},"cost":{"input":4.5455,"output":27.2727,"cache_read":0.4545},"canonical_model_id":"openai/gpt-5.5"},"openai/chatgpt-4o-latest":{"id":"openai/chatgpt-4o-latest","name":"ChatGPT-4o-Latest","description":"Legacy model retained for compatibility with older integrations","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-08-14","last_updated":"2024-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":4.5,"output":14}},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":1.8182,"output":7.2727,"cache_read":0.4545},"canonical_model_id":"openai/gpt-4.1"},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"o3-mini-high","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":0.99,"output":4},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.1,"output":9},"canonical_model_id":"openai/gpt-5-codex"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4-Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.6818,"output":4.0909,"cache_read":0.0682},"canonical_model_id":"openai/gpt-5.4-mini"},"lumalabs/ray2":{"id":"lumalabs/ray2","type":"video","name":"Ray2","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ray","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-02-20","last_updated":"2025-02-20","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":5000,"output":0}},"anthropic/claude-sonnet-3.5-june":{"id":"anthropic/claude-sonnet-3.5-june","name":"Claude-Sonnet-3.5-June","description":"Legacy model retained for compatibility with older integrations","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-11-18","last_updated":"2024-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2}},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude-Sonnet-4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":31999}],"tool_call":true,"temperature":false,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":32768},"cost":{"input":2.5758,"output":12.8788,"cache_read":0.2576,"cache_write":3.2197},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude-Opus-4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":128000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude-Opus-4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":0,"max":63999}],"tool_call":true,"temperature":false,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":64000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-3.7":{"id":"anthropic/claude-sonnet-3.7","name":"Claude-Sonnet-3.7","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-10-31","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":128000},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2},"canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"anthropic/claude-opus-4":{"id":"anthropic/claude-opus-4","name":"Claude-Opus-4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":192512,"output":28672},"status":"deprecated","cost":{"input":13,"output":64,"cache_read":1.3,"cache_write":16},"canonical_model_id":"anthropic/claude-opus-4-0"},"anthropic/claude-haiku-3":{"id":"anthropic/claude-haiku-3","name":"Claude-Haiku-3","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-08-31","release_date":"2024-03-13","last_updated":"2024-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":0.21,"output":1.1,"cache_read":0.021,"cache_write":0.26},"canonical_model_id":"anthropic/claude-3-haiku-20240307"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude-Opus-4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":31999}],"tool_call":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":32000},"status":"deprecated","cost":{"input":13,"output":64,"cache_read":1.3,"cache_write":16},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude-Haiku-4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":63999}],"tool_call":true,"temperature":false,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":192000,"output":64000},"cost":{"input":0.8586,"output":4.2929,"cache_read":0.0859,"cache_write":1.0732},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude-Sonnet-4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":128000},"cost":{"input":2.5758,"output":12.8788,"cache_read":0.2576,"cache_write":3.2197},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-4":{"id":"anthropic/claude-sonnet-4","name":"Claude-Sonnet-4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":64000},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"anthropic/claude-sonnet-3.5":{"id":"anthropic/claude-sonnet-3.5","name":"Claude-Sonnet-3.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2},"canonical_model_id":"anthropic/claude-3-5-sonnet-20241022"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude-Opus-4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":128000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude-Opus-4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":128000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-haiku-3.5":{"id":"anthropic/claude-haiku-3.5","name":"Claude-Haiku-3.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-07-31","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":0.68,"output":3.4,"cache_read":0.068,"cache_write":0.85},"canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"trytako/tako":{"id":"trytako/tako","name":"Tako","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"tako","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":0}},"novita/glm-5":{"id":"novita/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"novita/kimi-k2.5":{"id":"novita/kimi-k2.5","name":"Kimi-K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"status":"deprecated","cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"novita/glm-4.7-flash":{"id":"novita/glm-4.7-flash","name":"glm-4.7-flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":65500},"cost":{"input":0.0707,"output":0.404},"canonical_model_id":"zhipuai/glm-4.7-flash"},"novita/glm-4.7":{"id":"novita/glm-4.7","name":"glm-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"status":"deprecated","canonical_model_id":"zhipuai/glm-4.7"},"novita/minimax-m2.1":{"id":"novita/minimax-m2.1","name":"minimax-m2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"cost":{"input":0.303,"output":1.2121,"cache_read":0.1515},"canonical_model_id":"minimax/MiniMax-M2.1"},"novita/kimi-k2.6":{"id":"novita/kimi-k2.6","name":"Kimi-K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.9596,"output":4.0404,"cache_read":0.1616},"canonical_model_id":"moonshotai/kimi-k2.6"},"novita/glm-4.6":{"id":"novita/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"cost":{"input":0.6061,"output":2.2222,"cache_read":0.1111},"canonical_model_id":"zhipuai/glm-4.6"},"novita/glm-4.6v":{"id":"novita/glm-4.6v","name":"glm-4.6v","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":32768},"canonical_model_id":"zhipuai/glm-4.6v"},"novita/kimi-k2-thinking":{"id":"novita/kimi-k2-thinking","name":"kimi-k2-thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":0},"status":"deprecated","canonical_model_id":"moonshotai/kimi-k2-thinking"},"novita/glm-4.7-n":{"id":"novita/glm-4.7-n","name":"glm-4.7-n","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"canonical_model_id":"zhipuai/glm-4.7"},"novita/deepseek-v3.2":{"id":"novita/deepseek-v3.2","name":"DeepSeek-V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":0},"cost":{"input":0.2626,"output":0.3838,"cache_read":0.1313},"canonical_model_id":"deepseek/deepseek-v3.2"},"empiriolabs/deepseek-v4-pro-el":{"id":"empiriolabs/deepseek-v4-pro-el","name":"DeepSeek-V4-Pro-EL","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":1.67,"output":3.33},"canonical_model_id":"deepseek/deepseek-v4-pro"},"empiriolabs/deepseek-v4-flash-el":{"id":"empiriolabs/deepseek-v4-flash-el","name":"DeepSeek-V4-Flash-EL","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":0.14,"output":0.28},"canonical_model_id":"deepseek/deepseek-v4-flash"},"ideogramai/ideogram-v2a-turbo":{"id":"ideogramai/ideogram-v2a-turbo","type":"image","name":"Ideogram-v2a-Turbo","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-02-27","last_updated":"2025-02-27","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"ideogramai/ideogram-v2a":{"id":"ideogramai/ideogram-v2a","type":"image","name":"Ideogram-v2a","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-02-27","last_updated":"2025-02-27","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"ideogramai/ideogram-v2":{"id":"ideogramai/ideogram-v2","type":"image","name":"Ideogram-v2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-08-21","last_updated":"2024-08-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"ideogramai/ideogram":{"id":"ideogramai/ideogram","type":"image","name":"Ideogram","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-04-03","last_updated":"2024-04-03","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"fireworks-ai/kimi-k2.5-fw":{"id":"fireworks-ai/kimi-k2.5-fw","name":"Kimi-K2.5-FW","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":245760,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.5"}}},"tokenrouter":{"id":"tokenrouter","env":["TOKENROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tokenrouter.com/v1","name":"TokenRouter","doc":"https://www.tokenrouter.com/docs/tokenrouter-feature-guide/","models":{"z-ai/glm-5.3-free":{"id":"z-ai/glm-5.3-free","name":"GLM-5.3 (free)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"}}},"aimlapi":{"id":"aimlapi","env":["AIMLAPI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.aimlapi.com/v1","name":"AI/ML API","doc":"https://docs.aimlapi.com","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.9,"output":2.7,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-chat-v3-0324":{"id":"deepseek/deepseek-chat-v3-0324","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":124000},"cost":{"input":0.30682,"output":1.20612,"cache_read":0.14812},"canonical_model_id":"deepseek/deepseek-chat-v3-0324"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.84092,"output":3.68184,"cache_read":0.15341},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.39,"output":1.56,"cache_read":0.0078},"canonical_model_id":"deepseek/deepseek-v4-flash"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.116,"output":6.348,"cache_read":0.2645},"canonical_model_id":"sakana/fugu-max"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":31.74,"cache_read":0.529},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":31.74,"cache_read":0.529},"canonical_model_id":"sakana/fugu-ultra"},"inception/mercury-2":{"id":"inception/mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.2645,"output":0.7935,"cache_read":0.02645},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.2116,"output":0.7935,"cache_read":0.02116},"canonical_model_id":"inception/mercury-2.5"},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"Aion-RP 1.0 (8B)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.8464,"output":1.6928},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"V3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":81920,"output":38000},"cost":{"input":0.8464,"output":1.2696},"canonical_model_id":"morph/morph-v3-fast"},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"V3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.9522,"output":2.0102},"canonical_model_id":"morph/morph-v3-large"},"cognitivecomputations/dolphin-mistral-24b-venice-edition":{"id":"cognitivecomputations/dolphin-mistral-24b-venice-edition","name":"Venice Uncensored","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2116,"output":0.9522},"canonical_model_id":"cognitivecomputations/dolphin-mistral-24b-venice-edition"},"alibaba/qwen3.8-27b":{"id":"alibaba/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.4761,"output":3.3856},"canonical_model_id":"alibaba/qwen3.8-27b"},"alibaba/qwen3.8-max-prime":{"id":"alibaba/qwen3.8-max-prime","name":"Qwen3.8 Max Prime","description":"High-throughput edition of Qwen3.8 Max for coding, professional work, multimodal understanding, and long-running agent workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4.232,"output":12.696,"cache_read":0.529},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"alibaba/qwen3.8-2.4t-a95b":{"id":"alibaba/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":52429},"cost":{"input":2.645,"output":6.6125,"cache_read":0.529},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"alibaba/qwen3.8-flash":{"id":"alibaba/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"alibaba/qwen-plus":{"id":"alibaba/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131000,"output":16384},"cost":{"input":0.4,"output":1.2,"cache_read":0.08},"canonical_model_id":"alibaba/qwen-plus"},"alibaba/qwen-max":{"id":"alibaba/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":1.6,"output":6.4,"cache_read":0.32},"canonical_model_id":"alibaba/qwen-max"},"alibaba/qwen3-coder-480b-a35b-instruct":{"id":"alibaba/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":1.058,"output":5.15775,"cache_read":1.058},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"alibaba/qwen3-next-80b-a3b-instruct":{"id":"alibaba/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":129024,"output":16384},"cost":{"input":0.1587,"output":1.2696},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"x-ai/grok-4-20-0309-non-reasoning":{"id":"x-ai/grok-4-20-0309-non-reasoning","name":"Grok 4.20 Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1999000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"x-ai/grok-4-fast-non-reasoning":{"id":"x-ai/grok-4-fast-non-reasoning","name":"Grok 4 Fast Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1999000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.058,"output":4.2849,"cache_read":0.17986},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"cost":{"input":0.61364,"output":1.52352,"cache_read":0.122728},"canonical_model_id":"thinkingmachines/inkling-small"},"relace/relace-apply-3":{"id":"relace/relace-apply-3","name":"Apply 3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.8993,"output":1.3225},"canonical_model_id":"relace/relace-apply-3"},"relace/relace-search":{"id":"relace/relace-search","name":"Search","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":1.058,"output":3.174},"canonical_model_id":"relace/relace-search"},"sao10k/l3-lunaris-8b":{"id":"sao10k/l3-lunaris-8b","name":"Llama 3 8B Lunaris","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-13","last_updated":"2024-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":7372},"cost":{"input":0.0529,"output":0.0529},"canonical_model_id":"sao10k/l3-lunaris-8b"},"sao10k/l3.3-euryale-70b":{"id":"sao10k/l3.3-euryale-70b","name":"Llama 3.3 Euryale 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-12-18","last_updated":"2024-12-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.6877,"output":0.7935},"canonical_model_id":"sao10k/l3.3-euryale-70b"},"sao10k/l3.1-euryale-70b":{"id":"sao10k/l3.1-euryale-70b","name":"Llama 3.1 Euryale 70B v2.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1.534523,"output":1.534523},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"ibm-granite/granite-4.0-h-micro":{"id":"ibm-granite/granite-4.0-h-micro","name":"Granite 4.0 Micro","description":"Compact open-weight hybrid Granite model for lightweight enterprise chat and tool calling","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.017986,"output":0.118496},"canonical_model_id":"ibm/granite-4-h-micro"},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"output":4096},"cost":{"input":0.529,"output":0.7935},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":2.645,"output":10.58},"canonical_model_id":"cohere/command-a"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+ (08-2024)","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.645,"output":10.58},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B (12-2024)","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.039675,"output":0.1587},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R (08-2024)","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.1587,"output":0.6348},"canonical_model_id":"cohere/command-r-08-2024"},"bytedance-seed/seed-2.0-mini":{"id":"bytedance-seed/seed-2.0-mini","name":"Seed-2.0-Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1058,"output":0.4232},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance-seed/seed-1.6-flash":{"id":"bytedance-seed/seed-1.6-flash","name":"Seed 1.6 Flash","description":"Low-latency ByteDance Seed model for high-throughput chat, extraction, and lightweight tool use","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.07935,"output":0.3174},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"Seed-2.0-Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.2645,"output":2.116},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"Seed-2.0-Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.529,"output":3.174},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.14812,"output":0.29624,"cache_read":0.002962},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.46023,"output":0.92046,"cache_read":0.003809},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.03174,"output":0.12696,"cache_read":0.006348},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.1587,"output":0.6348,"cache_read":0.01587},"canonical_model_id":"upstage/solar-pro3"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":117964},"cost":{"input":0.1587,"output":0.48668,"cache_read":0.07935},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":96000},"cost":{"input":0.0529,"output":0.1587,"cache_read":0.04232},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.587,"output":9.522,"cache_read":0.1587,"cache_write":0.088167},"canonical_model_id":"google/gemini-flash-latest"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048756,"output":65536},"cost":{"input":2.116,"output":12.696,"cache_read":0.2116,"cache_write":0.39675},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemma-2-27b-it":{"id":"google/gemma-2-27b-it","name":"Gemma 2 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-07-13","last_updated":"2024-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0.6877,"output":0.6877},"canonical_model_id":"google/gemma-2-27b-it"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.1587,"output":0.6348,"cache_read":0.1587},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.2645,"output":1.587,"cache_read":0.02645,"cache_write":0.088167},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.5},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-8-flash":{"id":"google/gemini-3-8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":32768},"cost":{"input":0.0529,"output":0.1058,"cache_read":0.04232},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B Instruct","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":1.04742,"output":1.57642,"cache_read":1.04742},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1.058,"output":2.8566,"cache_read":0.0529},"canonical_model_id":"stepfun/step-5-preview"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":158.7,"output":634.8},"canonical_model_id":"openai/o1-pro"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra Pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.116,"output":12.696,"cache_read":0.2116},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":22.218,"output":177.744},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3 Mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":117964},"cost":{"input":0.07935,"output":0.3174,"cache_read":0.039675},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"gpt-oss-safeguard-20b","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.07935,"output":0.3174,"cache_read":0.039675},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":3.703,"output":29.624,"cache_read":0.3703},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o Mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":8192},"cost":{"input":30,"output":60,"cache_read":30},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1-Codex-Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2.645,"output":21.16,"cache_read":0.2645},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5-4":{"id":"openai/gpt-5-4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-1":{"id":"openai/gpt-5-1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5-5":{"id":"openai/gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":31.74,"output":190.44},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":31.74,"output":190.44},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15.87,"output":126.96},"canonical_model_id":"openai/gpt-5-pro"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-5-2":{"id":"openai/gpt-5-2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.232,"output":21.16,"cache_read":0.4232},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2116,"output":1.3225,"cache_read":0.02116},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":117964},"cost":{"input":0.3703,"output":1.0051,"cache_read":0.3703},"canonical_model_id":"openai/gpt-oss-120b"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3 Pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":21.16,"output":84.64},"canonical_model_id":"openai/o3-pro"},"openai/gpt-5.4-image-2":{"id":"openai/gpt-5.4-image-2","type":"image","name":"GPT-5.4 Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":8.464,"output":15.87,"cache_read":2.116},"canonical_model_id":"openai/gpt-5.4-image-2"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna Pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2116,"output":1.2696,"cache_read":0.02116},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.587,"output":9.522,"cache_read":0.1587},"canonical_model_id":"openai/gpt-5.4-mini"},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.06348,"output":0.25392},"canonical_model_id":"amazon/nova-lite-v1"},"amazon/nova-micro-v1":{"id":"amazon/nova-micro-v1","name":"Nova Micro 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":5120},"cost":{"input":0.03703,"output":0.14812},"canonical_model_id":"amazon/nova-micro-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.8464,"output":3.3856},"canonical_model_id":"amazon/nova-pro-v1"},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":100000},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.60306,"output":2.4334},"canonical_model_id":"moonshotai/kimi-k2"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":1.058,"output":3.174},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6348,"output":2.645,"cache_read":0.1587},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.3225,"output":4.4965,"cache_read":0.1587},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.3225,"output":4.4965,"cache_read":0.1587},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3703,"output":1.587,"cache_read":0.04232},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.3225,"output":4.4965,"cache_read":0.1587},"canonical_model_id":"meta/muse-spark-1.3"},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax 13B","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":1.98375,"output":1.98375,"cache_read":1.98375},"canonical_model_id":"gryphe/mythomax-l2-13b"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 2508","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":0.3174,"output":0.9522,"cache_read":0.03174},"canonical_model_id":"mistral/codestral-2508"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral NeMo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1587,"output":0.17986,"cache_read":0.1587},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-large-2407":{"id":"mistralai/mistral-large-2407","name":"Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2.116,"output":6.348,"cache_read":0.2116},"canonical_model_id":"mistral/mistral-large-2407"},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":26214},"cost":{"input":0.2116,"output":0.6348,"cache_read":0.02116},"canonical_model_id":"mistral/mistral-saba"},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.3703,"output":0.3703,"cache_read":0.02116},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2 2512","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.46552,"output":2.3276,"cache_read":0.046552},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4232,"output":2.116,"cache_read":0.04232},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1587,"output":0.1587,"cache_read":0.01058},"canonical_model_id":"mistral/ministral-3b-2512"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":102400},"cost":{"input":2.116,"output":6.348,"cache_read":0.2116},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-medium-3-5":{"id":"mistralai/mistral-medium-3-5","name":"Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.587,"output":7.935},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Small 3.2 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07935,"output":0.2116},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4232,"output":2.116,"cache_read":0.04232},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.0529,"output":0.08464},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Small 3.1 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.371358,"output":0.58719},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.198375,"output":0.7935,"cache_read":0.01587},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2.116,"output":6.348,"cache_read":0.2116},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.3174,"output":0.3174,"cache_read":0.01587},"canonical_model_id":"mistral/ministral-8b-2512"},"mistralai/voxtral-small-24b-2507":{"id":"mistralai/voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32768},"cost":{"input":0.1058,"output":0.3174,"cache_read":0.01058},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"thedrummer/skyfall-36b-v2":{"id":"thedrummer/skyfall-36b-v2","name":"Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.5819,"output":0.8464,"cache_read":0.2645},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"thedrummer/unslopnemo-12b":{"id":"thedrummer/unslopnemo-12b","name":"UnslopNemo 12B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-11-08","last_updated":"2024-11-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.4232,"output":0.4232},"canonical_model_id":"thedrummer/unslopnemo-12b"},"thedrummer/cydonia-24b-v4.1":{"id":"thedrummer/cydonia-24b-v4.1","name":"Cydonia 24B V4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3174,"output":0.529,"cache_read":0.1587},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"nousresearch/hermes-3-llama-3.1-405b":{"id":"nousresearch/hermes-3-llama-3.1-405b","name":"Hermes 3 405B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1.058,"output":1.058},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"nousresearch/hermes-3-llama-3.1-70b":{"id":"nousresearch/hermes-3-llama-3.1-70b","name":"Hermes 3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.7406,"output":0.7406},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"mancer/weaver":{"id":"mancer/weaver","name":"Weaver (alpha)","description":"General-purpose chat model for instruction following, writing, and analysis","family":"alpha","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-08-02","last_updated":"2023-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":2000},"cost":{"input":0.7935,"output":1.058},"canonical_model_id":"mancer/weaver"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.2116,"output":0.8464,"cache_read":0.0529},"canonical_model_id":"tencent/hy3"},"tencent/hy-mt2-7b":{"id":"tencent/hy-mt2-7b","name":"Hy-MT2-7B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.078292,"output":0.31211},"canonical_model_id":"tencent/hy-mt2-7b"},"tencent/hy-mt2-1.8b":{"id":"tencent/hy-mt2-1.8b","name":"Hy-MT2-1.8B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.046552,"output":0.187266},"canonical_model_id":"tencent/hy-mt2-1.8b"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.882372,"output":2.646058,"cache_read":0.044436},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy-mt2-30b-a3b":{"id":"tencent/hy-mt2-30b-a3b","name":"Hy-MT2-30B-A3B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.078292,"output":0.31211},"canonical_model_id":"tencent/hy-mt2-30b-a3b"},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03174,"output":0.1587},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.0529,"output":0.24334},"canonical_model_id":"inference-net/schematron-v2-small"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude 4.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.174,"output":15.87,"cache_read":0.3174,"cache_write":3.9675},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude 4.5 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.29,"output":26.45,"cache_read":0.529,"cache_write":6.6125},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude 4.5 Haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.058,"output":5.29,"cache_read":0.1058,"cache_write":1.3225},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude 4.6 Sonnet","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.174,"output":15.87,"cache_read":0.3174,"cache_write":3.9675},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude 4.7 Opus","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":26.45,"cache_read":0.529,"cache_write":6.6125},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude 4.8 Opus","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":26.45,"cache_read":0.529,"cache_write":6.6125},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6348,"output":1.9044,"cache_read":0.11638},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"cost":{"input":0.13225,"output":0.529,"cache_read":0.01058},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM 4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3174,"output":0.9522,"cache_read":0.05819},"canonical_model_id":"zhipuai/glm-4.6v"},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":16384},"cost":{"input":3.174,"output":5.29},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"meta-llama/Llama-3.3-70B-Instruct-Turbo":{"id":"meta-llama/Llama-3.3-70B-Instruct-Turbo","name":"Llama 3.3 70B Instruct Turbo","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":127000},"cost":{"input":1.04,"output":1.04},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"zhipu/glm-5.3":{"id":"zhipu/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zhipu/glm-5.2":{"id":"zhipu/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"unbiased/pareto":{"id":"unbiased/pareto","name":"Pareto","description":"Blended multimodal model that runs multiple language models in parallel and returns a synthesized answer","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.645,"output":7.935,"cache_read":0.2645},"canonical_model_id":"unbiased/pareto"},"rekaai/reka-edge":{"id":"rekaai/reka-edge","name":"Reka Edge","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.1058,"output":0.1058},"canonical_model_id":"rekaai/reka-edge"}}},"lowrouter":{"id":"lowrouter","env":["LOWROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lowrouter.ai/v1","name":"LowRouter","doc":"https://lowrouter.ai/docs/models/available","models":{"auto/openai/gpt-5.6-terra":{"id":"auto/openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"auto/openai/gpt-oss-safeguard-120b":{"id":"auto/openai/gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.18,"output":0.73},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"auto/openai/gpt-6.1-sol":{"id":"auto/openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"auto/openai/gpt-oss-20b":{"id":"auto/openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.113,"output":0.4745},"canonical_model_id":"openai/gpt-oss-20b"},"auto/openai/gpt-oss-safeguard-20b":{"id":"auto/openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.24},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"auto/openai/gpt-6-luna":{"id":"auto/openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"auto/openai/gpt-5.4":{"id":"auto/openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"auto/openai/gpt-5.6-luna":{"id":"auto/openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"auto/openai/gpt-4o-mini":{"id":"auto/openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"auto/openai/gpt-5.2":{"id":"auto/openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"auto/openai/gpt-4.1-mini":{"id":"auto/openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"auto/openai/gpt-5-mini":{"id":"auto/openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"auto/openai/gpt-6-astra":{"id":"auto/openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"auto/openai/gpt-4o":{"id":"auto/openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"auto/openai/gpt-oss-120b":{"id":"auto/openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1695,"output":0.6779},"canonical_model_id":"openai/gpt-oss-120b"},"auto/openai/gpt-5":{"id":"auto/openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"auto/openai/gpt-5-nano":{"id":"auto/openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"auto/openai/o3":{"id":"auto/openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"auto/openai/gpt-5.5":{"id":"auto/openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"auto/openai/gpt-5.6-sol":{"id":"auto/openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"auto/openai/gpt-6-sol":{"id":"auto/openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"auto/openai/gpt-4.1":{"id":"auto/openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"auto/openai/gpt-5.4-mini":{"id":"auto/openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"auto/mistralai/codestral-2508":{"id":"auto/mistralai/codestral-2508","name":"Codestral 25.08","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.3138,"output":0.9413,"cache_read":0.0314},"canonical_model_id":"mistral/codestral-2508"},"auto/mistralai/mistral-large-2512":{"id":"auto/mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5229,"output":1.5688,"cache_read":0.0523},"canonical_model_id":"mistral/mistral-large-2512"},"auto/mistralai/ministral-14b-2512":{"id":"auto/mistralai/ministral-14b-2512","name":"Ministral 3 14B","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2092,"output":0.2092,"cache_read":0.0209},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"auto/mistralai/ministral-3b-2512":{"id":"auto/mistralai/ministral-3b-2512","name":"Ministral 3 3B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1046,"output":0.1046,"cache_read":0.0105},"canonical_model_id":"mistral/ministral-3-3b-instruct-2512"},"auto/mistralai/mistral-medium-3.5":{"id":"auto/mistralai/mistral-medium-3.5","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5688,"output":7.8442,"cache_read":0.1569},"canonical_model_id":"mistral/mistral-medium-2604"},"auto/mistralai/mistral-large-4":{"id":"auto/mistralai/mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"status":"beta","cost":{"input":1.4367,"output":4.4156,"cache_read":0.1479},"canonical_model_id":"mistral/mistral-large-4"},"auto/mistralai/mistral-small-2603":{"id":"auto/mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.1569,"output":0.6275,"cache_read":0.0157},"canonical_model_id":"mistral/mistral-small-2603"},"auto/mistralai/ministral-8b-2512":{"id":"auto/mistralai/ministral-8b-2512","name":"Ministral 3 8B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1569,"output":0.1569,"cache_read":0.0157},"canonical_model_id":"mistral/ministral-3-8b-instruct-2512"}}},"tencent-tokenhub":{"id":"tencent-tokenhub","env":["TENCENT_TOKENHUB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://tokenhub.tencentmaas.com/v1","name":"Tencent TokenHub","doc":"https://cloud.tencent.com/document/product/1823/130050","models":{"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"tencent/hy3"},"hy3-preview":{"id":"hy3-preview","name":"Hy3 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"tencent/hy3-preview"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"}}},"impossibl":{"id":"impossibl","env":["IMPOSSIBL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.impossibl.com/v1","name":"Impossibl","doc":"https://impossibl.com/docs/models","models":{"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.19,"output":0.51,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-4.5-air"},"groq/gpt-oss-20b":{"id":"groq/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"groq/gpt-oss-120b":{"id":"groq/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.003},"canonical_model_id":"xiaomi/mimo-v2.5"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"cache_read":3,"tiers":[{"input":60,"output":270,"cache_read":3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270,"cache_read":3}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"cache_read":3,"tiers":[{"input":60,"output":270,"cache_read":3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270,"cache_read":3}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"qwen/qwen3.8-max-preview":{"id":"qwen/qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":2.5,"output":7.5},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"tier":{"type":"context","size":262144}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.05,"tiers":[{"input":1,"output":4,"cache_read":0.2,"tier":{"type":"context","size":262144}}],"context_over_200k":{"input":1,"output":4,"cache_read":0.2}},"canonical_model_id":"alibaba/qwen3.6-flash"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"fireworks/gpt-oss-20b":{"id":"fireworks/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.07,"output":0.3,"cache_read":0.035},"canonical_model_id":"openai/gpt-oss-20b"},"fireworks/glm-5.2":{"id":"fireworks/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"fireworks/gpt-oss-120b":{"id":"fireworks/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"}}},"aiand":{"id":"aiand","env":["AIAND_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.aiand.com/v1","name":"ai&","doc":"https://docs.aiand.com/","models":{"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"google/gemma-4-31b-it"},"motif-technologies/motif-3":{"id":"motif-technologies/motif-3","name":"Motif 3","description":"Motif 3 is a large-scale, decoder-only Mixture-of-Experts (MoE) language model with 314 billion total parameters and 13.2 billion parameters activated per token.","family":"motif","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2,"cache_read":0.2},"canonical_model_id":"motif-technologies/motif-3"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.08},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/deepseek-v4.1-flash":{"id":"deepseek-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":0.6,"cache_read":0.02},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/deepseek-v4-pro":{"id":"deepseek-ai/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1,"output":2.5,"cache_read":0.25},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/deepseek-v4-flash":{"id":"deepseek-ai/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.25,"cache_read":0.08},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":12.5,"cache_read":0.5},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.75,"output":3.5,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":3,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.32,"output":3.2,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"zai-org/glm-5.3":{"id":"zai-org/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":4,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":4,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-5.3-flash":{"id":"zai-org/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048550,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"openrouter":{"id":"openrouter","env":["OPENROUTER_API_KEY"],"npm":"@openrouter/ai-sdk-provider","api":"https://openrouter.ai/api/v1","name":"OpenRouter","doc":"https://openrouter.ai/models","models":{"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":80000},"cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.02,"output":1,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.672336,"output":1.344672,"cache_read":0.056028},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-chat-v3-0324":{"id":"deepseek/deepseek-chat-v3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":115200},"cost":{"input":0.29,"output":1.14,"cache_read":0.11},"canonical_model_id":"deepseek/deepseek-chat-v3-0324"},"deepseek/deepseek-chat-v3.1":{"id":"deepseek/deepseek-chat-v3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-chat-v3.1"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.091,"output":0.182,"cache_read":0.0182},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.00686},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-v3.1-terminus":{"id":"deepseek/deepseek-v3.1-terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":1}},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.04444,"output":0.13332,"cache_read":0.001414},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16000},"cost":{"input":0.2574,"output":1.0287},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":0.41}},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v3.2"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/sakana-namazu":{"id":"sakana/sakana-namazu","name":"Sakana Namazu","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"inception/mercury-3":{"id":"inception/mercury-3","name":"Mercury 3","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01}},"inception/mercury-2":{"id":"inception/mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.04,"output":0.15,"cache_read":0.004},"canonical_model_id":"inception/mercury-2.5"},"scaledown/summarize":{"id":"scaledown/summarize","name":"Summarize","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":4000},"cost":{"input":0.04,"output":0}},"scaledown/extract":{"id":"scaledown/extract","name":"Extract","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":0.04,"output":0}},"~x-ai/grok-latest":{"id":"~x-ai/grok-latest","name":"Grok Latest","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}}},"aion-labs/aion-3.5":{"id":"aion-labs/aion-3.5","name":"Aion 3.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75}},"aion-labs/aion-2.0":{"id":"aion-labs/aion-2.0","name":"Aion-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-21","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.8,"output":1.6,"cache_read":0.2},"canonical_model_id":"aion-labs/aion-2.0"},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"Aion-RP 1.0 (8B)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.8,"output":1.6},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"aion-labs/aion-3.0-mini":{"id":"aion-labs/aion-3.0-mini","name":"Aion-3.0-Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18},"canonical_model_id":"aion-labs/aion-3.0-mini"},"aion-labs/aion-3.0":{"id":"aion-labs/aion-3.0","name":"Aion-3.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75},"canonical_model_id":"aion-labs/aion-3.0"},"aion-labs/aion-3.5-mini":{"id":"aion-labs/aion-3.5-mini","name":"Aion 3.5 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18}},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"Morph V3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.8,"output":1.2},"canonical_model_id":"morph/morph-v3-fast"},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"Morph V3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.9,"output":1.9},"canonical_model_id":"morph/morph-v3-large"},"moondream/moondream3-preview":{"id":"moondream/moondream3-preview","name":"Moondream 3 Preview","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"moondream/moondream3.1-9b-a2b":{"id":"moondream/moondream3.1-9b-a2b","name":"Moondream 3.1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"reflection/beam-1.0-501b-a23b":{"id":"reflection/beam-1.0-501b-a23b","name":"Beam 1.0 501B A23B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.25,"output":1,"cache_read":0.05},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"cognitivecomputations/dolphin-mistral-24b-venice-edition":{"id":"cognitivecomputations/dolphin-mistral-24b-venice-edition","name":"Uncensored","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2,"output":0.9},"canonical_model_id":"cognitivecomputations/dolphin-mistral-24b-venice-edition"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.20-multi-agent":{"id":"x-ai/grok-4.20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":230400},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"thinkingmachines/inkling:free":{"id":"thinkingmachines/inkling:free","name":"Inkling (free)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":471859},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling-small:free":{"id":"thinkingmachines/inkling-small:free","name":"Inkling Small (free)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"thinkingmachines/inkling-small"},"relace/relace-apply-3":{"id":"relace/relace-apply-3","name":"Relace Apply 3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.85,"output":1.25},"canonical_model_id":"relace/relace-apply-3"},"relace/relace-search":{"id":"relace/relace-search","name":"Relace Search","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":1,"output":3},"canonical_model_id":"relace/relace-search"},"~z-ai/glm-flash-latest":{"id":"~z-ai/glm-flash-latest","name":"GLM Flash Latest","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.0304,"output":0.122,"cache_read":0.03}},"~z-ai/glm-latest":{"id":"~z-ai/glm-latest","name":"GLM Latest","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-19","last_updated":"2026-08-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.079,"output":1.4,"cache_read":0.078}},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.075,"output":0.225,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.1-flash"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.021,"output":0.063,"cache_read":0.0042},"canonical_model_id":"inclusionai/ling-3.0-flash"},"sao10k/l3-lunaris-8b":{"id":"sao10k/l3-lunaris-8b","name":"Llama 3 8B Lunaris","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-13","last_updated":"2024-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":7372},"cost":{"input":0.04,"output":0.05},"canonical_model_id":"sao10k/l3-lunaris-8b"},"sao10k/l3.1-euryale-70b":{"id":"sao10k/l3.1-euryale-70b","name":"Llama 3.1 Euryale 70B v2.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.85,"output":0.85},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":32768},"cost":{"input":0.27,"output":0.95,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-01":{"id":"minimax/minimax-01","name":"MiniMax-01","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-03-31","release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000192,"output":40000},"cost":{"input":0.2,"output":1.1}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.255,"output":1.02},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m1":{"id":"minimax/minimax-m1","name":"MiniMax M1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":40000},"cost":{"input":0.55,"output":2.2}},"minimax/minimax-m2-her":{"id":"minimax/minimax-m2-her","name":"MiniMax-M2 Her","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":2048},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2-Her"},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"ibm-granite/granite-4.0-h-micro":{"id":"ibm-granite/granite-4.0-h-micro","name":"Granite 4.0 Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"granite","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":117900},"cost":{"input":0.017,"output":0.112}},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"output":5529},"cost":{"input":0.35,"output":0.65},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-a-plus":{"id":"cohere/command-a-plus","name":"Command A+","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04-01","release_date":"2026-05-20","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":64000},"cost":{"input":0.3,"output":1.5,"cache_read":0.15},"canonical_model_id":"cohere/command-a-plus-05-2026"},"cohere/north-mini-code:free":{"id":"cohere/north-mini-code:free","name":"North Mini Code (free)","description":"Cohere coding model for practical software engineering and agentic edits","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"bytedance-seed/seed-1.6":{"id":"bytedance-seed/seed-1.6","name":"Seed 1.6","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.25,"output":2,"tiers":[{"input":0.5,"output":4,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"bytedance-seed/seed-2.0-mini":{"id":"bytedance-seed/seed-2.0-mini","name":"Seed 2.0 Mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.4,"tiers":[{"input":0.2,"output":0.8,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance-seed/seed-1.6-flash":{"id":"bytedance-seed/seed-1.6-flash","name":"Seed 1.6 Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.075,"output":0.3,"tiers":[{"input":0.1,"output":0.8,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"bytedance-seed/seed-2-1-turbo":{"id":"bytedance-seed/seed-2-1-turbo","name":"Seed 2.1 Turbo","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":2.5},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"Seed 2.0 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.25,"output":2,"tiers":[{"input":0.5,"output":4,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.5,"output":3,"tiers":[{"input":1,"output":6,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"microsoft/phi-4":{"id":"microsoft/phi-4","name":"Phi 4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-01-10","last_updated":"2025-01-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":14745},"cost":{"input":0.07,"output":0.14}},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.09,"output":0.18,"cache_read":0.009},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-xs-2.1:free":{"id":"poolside/laguna-xs-2.1:free","name":"Laguna XS 2.1 (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.06,"output":0.12,"cache_read":0.03},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-s-2.1:free":{"id":"poolside/laguna-s-2.1:free","name":"Laguna S 2.1 (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-s-2.1"},"~anthropic/claude-fable-latest":{"id":"~anthropic/claude-fable-latest","name":"Claude Fable Latest","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5}},"~anthropic/claude-sonnet-latest":{"id":"~anthropic/claude-sonnet-latest","name":"Claude Sonnet Latest","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"~anthropic/claude-haiku-latest":{"id":"~anthropic/claude-haiku-latest","name":"Claude Haiku Latest","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]}},"~anthropic/claude-opus-latest":{"id":"~anthropic/claude-opus-latest","name":"Claude Opus Latest","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5}},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.119,"output":0.238,"cache_read":0.00255},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.3045,"output":0.609,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"perceptron/perceptron-mk1":{"id":"perceptron/perceptron-mk1","name":"Perceptron Mk1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"perceptron/perceptron-mk1"},"perceptron/perceptron-mk1.5":{"id":"perceptron/perceptron-mk1.5","name":"Perceptron Mk1.5","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":36864,"output":8192},"cost":{"input":0.15,"output":1.5}},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.15,"output":0.6,"cache_read":0.03},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":117964},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-mini4":{"id":"upstage/solar-mini4","name":"Solar Mini 4","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.005},"canonical_model_id":"upstage/solar-mini4"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.08,"output":0.35,"cache_read":0.04},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"reasoning":10,"cache_read":0.125,"cache_write":0.375,"input_audio":1.25,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/lyria-3-clip-preview":{"id":"google/lyria-3-clip-preview","name":"Lyria 3 Clip Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-clip-preview"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemma-2-27b-it":{"id":"google/gemma-2-27b-it","name":"Gemma 2 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-07-13","last_updated":"2024-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0.65,"output":0.65},"canonical_model_id":"google/gemma-2-27b-it"},"google/gemma-4-26b-a4b-it:free":{"id":"google/gemma-4-26b-a4b-it:free","name":"Gemma 4 26B A4B (free)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.0675,"output":0.225,"cache_read":0.0375},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/lyria-3-pro-preview":{"id":"google/lyria-3-pro-preview","name":"Lyria 3 Pro Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-pro-preview"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"reasoning":3,"cache_read":0.05,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemini-2.5-pro-preview":{"id":"google/gemini-2.5-pro-preview","name":"Gemini 2.5 Pro Preview 06-05","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["pdf","image","text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"reasoning":10,"cache_read":0.125,"cache_write":0.375,"input_audio":1.25,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro-preview"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.04,"output":0.05,"cache_read":0.02},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.825,"output":4.125,"reasoning":4.125,"cache_read":0.0825,"cache_write":0.041667,"input_audio":0.825},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemma-4-31b-it:free":{"id":"google/gemma-4-31b-it:free","name":"Gemma 4 31B (free)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.1,"output":0.4,"reasoning":0.4,"cache_read":0.01,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":256000,"output":230400},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1000000,"output":64000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":150,"output":600},"canonical_model_id":"openai/o1-pro"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-audio-mini":{"id":"openai/gpt-audio-mini","type":"chat","name":"GPT-Audio Mini","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":2.4,"input_audio":0.6,"output_audio":2.4},"canonical_model_id":"openai/gpt-audio-mini"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6.1-sol-pro":{"id":"openai/gpt-6.1-sol-pro","name":"GPT-6.1 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}}},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.2-chat":{"id":"openai/gpt-5.2-chat","name":"GPT-5.2 Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.018,"output":0.09,"cache_read":0.009},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-3.5-turbo-16k":{"id":"openai/gpt-3.5-turbo-16k","name":"GPT-3.5 Turbo 16k","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-30","release_date":"2023-08-28","last_updated":"2023-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":3,"output":4}},"openai/gpt-chat-latest":{"id":"openai/gpt-chat-latest","name":"GPT Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5}},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":4096},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/o4-mini-high":{"id":"openai/o4-mini-high","name":"o4 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-06-30","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275}},"openai/gpt-3.5-turbo-instruct":{"id":"openai/gpt-3.5-turbo-instruct","name":"GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2021-09-30","release_date":"2023-09-28","last_updated":"2023-09-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1.5,"output":2}},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra-pro":{"id":"openai/gpt-6-astra-pro","name":"GPT-6 Astra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}}},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5-image-mini":{"id":"openai/gpt-5-image-mini","type":"image","name":"GPT-5 Image Mini","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["pdf","image","text"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":2,"cache_read":0.25},"canonical_model_id":"openai/gpt-5-image-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-4o-mini-2024-07-18":{"id":"openai/gpt-4o-mini-2024-07-18","name":"GPT-4o-mini (2024-07-18)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-6-sol-pro":{"id":"openai/gpt-6-sol-pro","name":"GPT-6 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}}},"openai/gpt-6-luna-pro":{"id":"openai/gpt-6-luna-pro","name":"GPT-6 Luna Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}}},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-audio":{"id":"openai/gpt-audio","type":"chat","name":"GPT-Audio","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-audio"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","pdf","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-5.4-image-2":{"id":"openai/gpt-5.4-image-2","type":"image","name":"GPT-5.4 Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":8,"output":15,"cache_read":2},"canonical_model_id":"openai/gpt-5.4-image-2"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5-image":{"id":"openai/gpt-5-image","type":"image","name":"GPT-5 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-10-01","release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":10,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-5-image"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-3.5-turbo-0613":{"id":"openai/gpt-3.5-turbo-0613","name":"GPT-3.5 Turbo (older v0613)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-30","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1,"output":2}},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"o3 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-10-31","release_date":"2025-02-12","last_updated":"2025-02-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"amazon/nova-premier-v1":{"id":"amazon/nova-premier-v1","name":"Nova Premier 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-30","last_updated":"2025-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"status":"deprecated","cost":{"input":2.5,"output":12.5,"cache_read":0.625},"canonical_model_id":"amazon/nova-premier"},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"amazon/nova-lite-v1"},"amazon/nova-2-lite-v1":{"id":"amazon/nova-2-lite-v1","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"amazon/nova-2-lite-v1"},"amazon/nova-micro-v1":{"id":"amazon/nova-micro-v1","name":"Nova Micro 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":5120},"cost":{"input":0.035,"output":0.14},"canonical_model_id":"amazon/nova-micro-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.8,"output":3.2},"canonical_model_id":"amazon/nova-pro-v1"},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Sonar Pro","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"perplexity/sonar-reasoning-pro":{"id":"perplexity/sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded reasoning model for multi-step research and cited answers","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8}},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-27","last_updated":"2025-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127072,"output":114364},"cost":{"input":1,"output":1}},"perplexity/sonar-deep-research":{"id":"perplexity/sonar-deep-research","name":"Sonar Deep Research","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar-deep-research","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8,"reasoning":3}},"perplexity/sonar-pro-search":{"id":"perplexity/sonar-pro-search","name":"Sonar Pro Search","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-30","last_updated":"2025-10-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"bytedance/ui-tars-1.5-7b":{"id":"bytedance/ui-tars-1.5-7b","name":"UI-TARS 7B ","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"cost":{"input":0.1,"output":0.2,"cache_read":0.1}},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.239,"output":10.8,"cache_read":0.21},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.65,"output":3.41,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 Instruct 0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-wonder":{"id":"meta/muse-wonder","name":"Muse Wonder","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":4,"output":20,"cache_read":0.2}},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-08-21","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax L2 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":3686},"cost":{"input":0.08,"output":0.11},"canonical_model_id":"gryphe/mythomax-l2-13b"},"~google/gemini-flash-latest":{"id":"~google/gemini-flash-latest","name":"Gemini Flash Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75}},"~google/gemini-pro-latest":{"id":"~google/gemini-pro-latest","name":"Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["audio","pdf","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}}},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 25.08","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":204800},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral Nemo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.029,"output":0.03},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-large-2407":{"id":"mistralai/mistral-large-2407","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-2407"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":26214},"cost":{"input":0.2,"output":0.6,"cache_read":0.02},"canonical_model_id":"mistral/mistral-saba"},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3b-2512"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Mistral Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":102400},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-medium-3-5":{"id":"mistralai/mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistralai/mistral-large-4-0":{"id":"mistralai/mistral-large-4-0","name":"Mistral Large 4","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Mistral Small 3.2 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.075,"output":0.2},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Mistral Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":102400},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-8b-2512"},"mistralai/voxtral-small-24b-2507":{"id":"mistralai/voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":26214},"cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":235929},"cost":{"input":0.2,"output":2.5,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen2.5-vl-72b-instruct":{"id":"qwen/qwen2.5-vl-72b-instruct","name":"Qwen2.5-VL 72B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":115200},"cost":{"input":0.8,"output":1,"cache_read":0.4},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"qwen/qwen3-235b-a22b-2507":{"id":"qwen/qwen3-235b-a22b-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.5-plus-20260420":{"id":"qwen/qwen3.5-plus-20260420","name":"Qwen3.5 Plus 2026-04-20","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.8,"cache_write":0.375,"tiers":[{"input":0.375,"output":2.25,"cache_write":0.46875,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.375,"output":2.25,"cache_write":0.46875}}},"qwen/qwen3.8-max-prime":{"id":"qwen/qwen3.8-max-prime","name":"Qwen 3.8 Max Prime","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"qwen/qwen3-vl-30b-a3b-thinking":{"id":"qwen/qwen3-vl-30b-a3b-thinking","name":"Qwen3 VL 30B A3B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.29,"output":1}},"qwen/qwen3.5-plus-02-15":{"id":"qwen/qwen3.5-plus-02-15","name":"Qwen3.5 Plus 2026-02-15","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.26,"output":1.56,"tiers":[{"input":0.325,"output":1.95,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.325,"output":1.95}}},"qwen/qwen3.5-flash-02-23":{"id":"qwen/qwen3.5-flash-02-23","name":"Qwen3.5-Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-25","last_updated":"2026-02-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.065,"output":0.26}},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.21,"output":1.9,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"cache_write":0.40625,"tiers":[{"input":1.3,"output":3.9,"cache_write":1.625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.3,"output":3.9,"cache_write":1.625}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-coder-30b-a3b-instruct":{"id":"qwen/qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.195,"output":0.975,"cache_read":0.039,"cache_write":0.24375,"tiers":[{"input":0.325,"output":1.625,"cache_read":0.065,"cache_write":0.40625,"tier":{"type":"context","size":32000}},{"input":0.52,"output":2.6,"cache_read":0.104,"cache_write":0.65,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-vl-8b-instruct":{"id":"qwen/qwen3-vl-8b-instruct","name":"Qwen3 VL 8B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.25,"output":0.75,"cache_read":0.12}},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"status":"deprecated","cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B-A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.45,"output":3.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B-A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.09,"output":0.3},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":3.6,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen-2.5-72b-instruct":{"id":"qwen/qwen-2.5-72b-instruct","name":"Qwen2.5 72B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.36,"output":0.4}},"qwen/qwen-2.5-7b-instruct":{"id":"qwen/qwen-2.5-7b-instruct","name":"Qwen2.5 7B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-16","last_updated":"2024-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.1,"output":0.2}},"qwen/qwen3-14b":{"id":"qwen/qwen3-14b","name":"Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.24}},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":909000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.26,"output":0.78,"cache_read":0.052,"cache_write":0.325,"tiers":[{"input":0.78,"output":2.34,"cache_read":0.156,"cache_write":0.975,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.78,"output":2.34,"cache_read":0.156,"cache_write":0.975}},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.32,"output":1.28,"cache_read":0.064,"cache_write":0.4,"tiers":[{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen-2.5-coder-32b-instruct":{"id":"qwen/qwen-2.5-coder-32b-instruct","name":"Qwen2.5-Coder-32B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.27,"output":2.16,"cache_read":0.27},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.12,"output":0.8,"cache_read":0.07},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.1,"output":0.95,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1875,"output":1.125,"cache_write":0.234375,"tiers":[{"input":0.75,"output":3,"cache_write":0.9375,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.75,"output":3,"cache_write":0.9375}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3-coder":{"id":"qwen/qwen3-coder","name":"Qwen3 Coder 480B A35B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":1,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"~deepseek/deepseek-v4-flash-latest":{"id":"~deepseek/deepseek-v4-flash-latest","name":"DeepSeek V4 Flash Latest","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-01","last_updated":"2026-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.0422,"output":0.1729,"cache_read":0.027}},"~deepseek/deepseek-flash-latest":{"id":"~deepseek/deepseek-flash-latest","name":"DeepSeek Flash Latest","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.02,"output":1,"cache_read":0.015}},"~deepseek/deepseek-pro-latest":{"id":"~deepseek/deepseek-pro-latest","name":"DeepSeek Pro Latest","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":393216},"cost":{"input":0.1188,"output":8,"cache_read":0.12}},"elorian/glimpse":{"id":"elorian/glimpse","name":"Glimpse","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":256000},"cost":{"input":0.6,"output":3,"cache_read":0.06}},"openrouter/auto":{"id":"openrouter/auto","name":"Auto Router","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2023-11-08","last_updated":"2023-11-08","modalities":{"input":["text","image","audio","pdf","video"],"output":["text","image"]},"open_weights":false,"limit":{"context":2000000,"output":2000000}},"openrouter/bodybuilder":{"id":"openrouter/bodybuilder","name":"Body Builder (beta)","description":"Preview model for early access evaluation, prototyping, and compatibility testing","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-12-05","last_updated":"2025-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000}},"openrouter/fusion":{"id":"openrouter/fusion","name":"Fusion","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"openrouter/pareto-code":{"id":"openrouter/pareto-code","name":"Pareto Code Router","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":200000}},"openrouter/free":{"id":"openrouter/free","name":"Free Models Router","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-01","last_updated":"2026-02-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":8000},"cost":{"input":0,"output":0}},"liquid/lfm-2.5-2.6b:free":{"id":"liquid/lfm-2.5-2.6b:free","name":"LFM2.5-2.6B (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"cost":{"input":0,"output":0}},"liquid/lfm-2.5-24b-a2b":{"id":"liquid/lfm-2.5-24b-a2b","name":"LFM2.5-24B-A2B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"thedrummer/skyfall-36b-v2":{"id":"thedrummer/skyfall-36b-v2","name":"Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.55,"output":0.8,"cache_read":0.25},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"thedrummer/unslopnemo-12b":{"id":"thedrummer/unslopnemo-12b","name":"UnslopNemo 12B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-11-08","last_updated":"2024-11-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":819200},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"thedrummer/unslopnemo-12b"},"thedrummer/cydonia-24b-v4.1":{"id":"thedrummer/cydonia-24b-v4.1","name":"Cydonia 24B V4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.3,"output":0.5,"cache_read":0.15},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"writer/palmyra-x5":{"id":"writer/palmyra-x5","name":"Palmyra X5","description":"General-purpose chat model for instruction following, writing, and analysis","family":"palmyra","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"output":8192},"cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"~moonshotai/kimi-latest":{"id":"~moonshotai/kimi-latest","name":"Kimi Latest","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.238,"output":24,"cache_read":0.228}},"nousresearch/hermes-3-llama-3.1-405b":{"id":"nousresearch/hermes-3-llama-3.1-405b","name":"Hermes 3 405B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":1,"output":1},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"nousresearch/hermes-3-llama-3.1-70b":{"id":"nousresearch/hermes-3-llama-3.1-70b","name":"Hermes 3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":0.7},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"nousresearch/hermes-4-405b":{"id":"nousresearch/hermes-4-405b","name":"Hermes 4 405B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":1,"output":3},"canonical_model_id":"nousresearch/hermes-4-405b"},"mancer/weaver":{"id":"mancer/weaver","name":"Weaver (alpha)","description":"General-purpose chat model for instruction following, writing, and analysis","family":"alpha","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-08-02","last_updated":"2023-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":6000},"cost":{"input":0.4,"output":0.75},"canonical_model_id":"mancer/weaver"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.0825,"output":0.33,"cache_read":0.020625},"canonical_model_id":"tencent/hy3"},"tencent/hy3-preview":{"id":"tencent/hy3-preview","name":"Hy3 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.18,"output":0.6,"cache_read":0.06},"canonical_model_id":"tencent/hy3-preview"},"tencent/hy-mt2-7b":{"id":"tencent/hy-mt2-7b","name":"Hy-MT2-7B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-7b"},"tencent/hy-mt2-1.8b":{"id":"tencent/hy-mt2-1.8b","name":"Hy-MT2-1.8B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.044,"output":0.177},"canonical_model_id":"tencent/hy-mt2-1.8b"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.7506,"output":2.2509,"cache_read":0.0378},"canonical_model_id":"tencent/hy4-preview"},"tencent/hunyuan-a13b-instruct":{"id":"tencent/hunyuan-a13b-instruct","name":"Hunyuan A13B Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.14,"output":0.57}},"tencent/hy-mt2-30b-a3b":{"id":"tencent/hy-mt2-30b-a3b","name":"Hy-MT2-30B-A3B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-30b-a3b"},"nex-agi/nex-n2.5-mini":{"id":"nex-agi/nex-n2.5-mini","name":"Nex-N2.5-Mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.025,"output":0.1,"cache_read":0.0025}},"nex-agi/nex-n2.5-pro":{"id":"nex-agi/nex-n2.5-pro","name":"Nex-N2.5-Pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.075,"output":0.25,"cache_read":0.015}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.05},"canonical_model_id":"inference-net/schematron-v2-small"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3-prime":{"id":"z-ai/glm-5.3-prime","name":"GLM 5.3 Prime","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.56}},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.079,"output":1.4,"cache_read":0.078},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM 5.3 FlashX","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.09}},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":117964},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.54,"output":1.98,"cache_read":0.099},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"status":"deprecated","cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.0341,"output":7,"cache_read":0.033},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":2.5,"output":5},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"meta-llama/llama-4-scout":{"id":"meta-llama/llama-4-scout","name":"Llama 4 Scout","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1310720,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama-3.1-8B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.05,"output":0.08,"cache_read":0.025},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-guard-4-12b":{"id":"meta-llama/llama-guard-4-12b","name":"Llama Guard 4 12B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"meta-llama/llama-3.2-3b-instruct":{"id":"meta-llama/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":72000},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"meta-llama/llama-3.1-70b-instruct":{"id":"meta-llama/llama-3.1-70b-instruct","name":"Llama-3.1-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"meta-llama/llama-4-maverick":{"id":"meta-llama/llama-4-maverick","name":"Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":16384},"cost":{"input":0.1875,"output":0.6525,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/llama-3.2-1b-instruct":{"id":"meta-llama/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":54000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"nvidia/nemotron-3.5-super-vl:free":{"id":"nvidia/nemotron-3.5-super-vl:free","name":"Nemotron 3.5 Super VL (free)","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3.5-content-safety:free":{"id":"nvidia/nemotron-3.5-content-safety:free","name":"Nemotron 3.5 Content Safety (free)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"Nemotron 3.5 Super VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":235929},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3-super-120b-a12b:free":{"id":"nvidia/nemotron-3-super-120b-a12b:free","name":"Nemotron 3 Super (free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nemotron 3.5 Lightning 30B A3B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.07,"output":0.2,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":182520},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free":{"id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free","name":"Nemotron 3 Nano Omni (free)","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3.5-lightning:free":{"id":"nvidia/nemotron-3.5-lightning:free","name":"Nemotron 3.5 Lightning (free)","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.3,"output":2.4,"cache_read":0.15},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b:free":{"id":"nvidia/nemotron-3-ultra-550b-a55b:free","name":"Nemotron 3 Ultra (free)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"~openai/gpt-mini-latest":{"id":"~openai/gpt-mini-latest","name":"GPT Mini Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075}},"~openai/gpt-astra-latest":{"id":"~openai/gpt-astra-latest","name":"GPT Astra Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}}},"~openai/gpt-luna-latest":{"id":"~openai/gpt-luna-latest","name":"GPT Luna Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}}},"~openai/gpt-sol-latest":{"id":"~openai/gpt-sol-latest","name":"GPT Sol Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}}},"~openai/gpt-terra-latest":{"id":"~openai/gpt-terra-latest","name":"GPT Terra Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}}},"meituan/longcat-2.0":{"id":"meituan/longcat-2.0","name":"LongCat-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048756,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.0"},"xingchen-agi/xing4.0-29b-a4b:free":{"id":"xingchen-agi/xing4.0-29b-a4b:free","name":"Xing4.0 29B A4B (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"agi","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"unbiased/pareto-26.10":{"id":"unbiased/pareto-26.10","name":"Pareto 26.10","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"unbiased/pareto-26.9":{"id":"unbiased/pareto-26.9","name":"Pareto 26.9","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25}},"prism-ml/ternary-bonsai-2-27b":{"id":"prism-ml/ternary-bonsai-2-27b","name":"Ternary Bonsai 2 27B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.5,"cache_read":0.0375}},"recursive/flint-v1.0:free":{"id":"recursive/flint-v1.0:free","name":"Flint (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"flint","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember-1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"dots-studio/dots-3-note-preview:free":{"id":"dots-studio/dots-3-note-preview:free","name":"Dots3-Note Preview (free)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":460800},"status":"deprecated","cost":{"input":0,"output":0}},"~unbiased/pareto-latest":{"id":"~unbiased/pareto-latest","name":"Pareto Latest","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"rekaai/reka-flash-3":{"id":"rekaai/reka-flash-3","name":"Reka Flash 3","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"reka","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":58982},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"rekaai/reka-flash-3"},"rekaai/reka-edge":{"id":"rekaai/reka-edge","name":"Reka Edge","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":14745},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"rekaai/reka-edge"}}},"agnes":{"id":"agnes","env":["AGNES_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://apihub.agnes-ai.com/v1","name":"Agnes AI","doc":"https://agnes-ai.com/doc","models":{"agnes-2.0-flash":{"id":"agnes-2.0-flash","name":"Agnes 2.0 Flash","description":"Fast and efficient model for agent workflows, tool calling, coding, and image understanding.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-25","last_updated":"2026-05-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":65536},"status":"deprecated","cost":{"input":0,"output":0}},"agnes-2.5-pro-alpha":{"id":"agnes-2.5-pro-alpha","name":"Agnes 2.5 Pro Alpha","description":"Paid reasoning model for advanced coding, scientific reasoning, long-context analysis, agentic workflows, and multimodal understanding.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0.45,"output":0.9,"cache_read":0.0038}},"agnes-3.0-flash":{"id":"agnes-3.0-flash","name":"Agnes 3.0 Flash","description":"A new-generation Agnes AI text model with stronger Agnes Code task execution, tool orchestration, and trustworthy delivery.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09","last_updated":"2026-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":65536},"cost":{"input":0,"output":0}},"agnes-2.5-flash":{"id":"agnes-2.5-flash","name":"Agnes 2.5 Flash","description":"Upgraded model with improved coding, agent workflows, tool calling, and multimodal understanding.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07","last_updated":"2026-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":65536},"cost":{"input":0,"output":0}}}},"vispark":{"id":"vispark","env":["VISPARK_LAB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lab.vispark.in/v1","name":"Vispark","doc":"https://lab.vispark.in/#vision","models":{"vispark/vision-large":{"id":"vispark/vision-large","name":"Vision Large","description":"Most capable Vision model for complex reasoning, detailed media analysis, and structured output over a 1M-token context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":7.37,"output":22.11},"canonical_model_id":"vispark/vision-large"},"vispark/vision-small":{"id":"vispark/vision-small","name":"Vision Small","description":"Fast, low-cost multimodal model for understanding text, images, audio, video, and PDFs, with tool calling and a 1M-token context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.05,"output":3.16},"canonical_model_id":"vispark/vision-small"},"vispark/vision-medium":{"id":"vispark/vision-medium","name":"Vision Medium","description":"Balanced multimodal model pairing a 1M-token context window with deeper reasoning for analysis, content creation, and tool use across text, image, audio, video, and PDF inputs.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":4.21,"output":12.63},"canonical_model_id":"vispark/vision-medium"}}},"alibaba-coding-plan":{"id":"alibaba-coding-plan","env":["ALIBABA_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://coding-intl.dashscope.aliyuncs.com/v1","name":"Alibaba Coding Plan","doc":"https://www.alibabacloud.com/help/en/model-studio/coding-plan","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-11","last_updated":"2026-02-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-max-2026-01-23":{"id":"qwen3-max-2026-01-23","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"input":196601,"output":24576},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}}}},"tokengo":{"id":"tokengo","env":["TOKENGO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tokengo.com/v1","name":"TokenGo","doc":"https://www.tokengo.com/docs","models":{"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.435,"output":0.87},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.098,"output":0.196,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.19,"output":0.71,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":0.2174,"output":0.326,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v3.2"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":2.65,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.89,"output":3.2647,"cache_read":0.2226},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"liquid":{"id":"liquid","env":["LIQUID_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://router.inference.ai.exchange/v1","name":"Liquid Inference","doc":"https://inference.ai.exchange/docs","models":{"liquid.auto":{"id":"liquid.auto","name":"Liquid Auto","description":"Routes each request to the lowest-priced eligible offer on the Liquid Inference exchange","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"output":1048576}}}},"github-copilot":{"id":"github-copilot","env":["GITHUB_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://api.githubcopilot.com","name":"GitHub Copilot","doc":"https://docs.github.com/en/copilot","models":{"claude-haiku-5.5":{"id":"claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":372000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]},{"type":"budget_tokens","min":256,"max":32000}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"claude-opus-5.5":{"id":"claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":224000,"output":32000},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"mai-code-1.1-flash":{"id":"mai-code-1.1-flash","name":"MAI-Code-1.1-Flash","description":"Microsoft coding model with native vision support, optimized for fast and efficient software development","family":"mai","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":128000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"microsoft/mai-code-1.1-flash"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-sonnet-5.5":{"id":"claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":264000,"input":128000,"output":64000},"status":"deprecated","cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"claude-haiku-4.5":{"id":"claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":136000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-sonnet-4.6":{"id":"claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":168000,"output":32000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":372000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"claude-opus-4.7":{"id":"claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":168000,"output":32000},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-opus-4.8":{"id":"claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":168000,"output":64000},"experimental":{"modes":{"fast":{"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-fable-5.1":{"id":"claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]},{"type":"budget_tokens","min":256,"max":24000}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":128000,"output":64000},"status":"deprecated","cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":372000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.5"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"helicone":{"id":"helicone","env":["HELICONE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ai-gateway.helicone.ai/v1","name":"Helicone","doc":"https://helicone.ai/models","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Anthropic: Claude 4.5 Haiku (20251001)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"sonar-pro":{"id":"sonar-pro","name":"Perplexity Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":4096},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"grok-3-mini":{"id":"grok-3-mini","name":"xAI Grok 3 Mini","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":0.5,"cache_read":0.075}},"llama-prompt-guard-2-86m":{"id":"llama-prompt-guard-2-86m","name":"Meta Llama Prompt Guard 2 86M","description":"86M-parameter (mDeBERTa-base) multilingual classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":2},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"meta/llama-prompt-guard-2-86m"},"ernie-4.5-21b-a3b-thinking":{"id":"ernie-4.5-21b-a3b-thinking","name":"Baidu Ernie 4.5 21B A3B Thinking","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"ernie","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-03","release_date":"2025-03-16","last_updated":"2025-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8000},"cost":{"input":0.07,"output":0.28}},"llama-4-scout":{"id":"llama-4-scout","name":"Meta Llama 4 Scout 17B 16E","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.08,"output":0.3},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"llama-3.3-70b-versatile":{"id":"llama-3.3-70b-versatile","name":"Meta Llama 3.3 70B Versatile","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32678},"cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"OpenAI: GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"gemma-3-12b-it":{"id":"gemma-3-12b-it","name":"Google Gemma 3 12B","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-12","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-12b-it"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Meta Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16400},"cost":{"input":0.13,"output":0.39},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"llama-3.1-8b-instruct":{"id":"llama-3.1-8b-instruct","name":"Meta Llama 3.1 8B Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.02,"output":0.05},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"llama-guard-4":{"id":"llama-guard-4","name":"Meta Llama Guard 4 12B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":1024},"cost":{"input":0.21,"output":0.21}},"qwen3-vl-235b-a22b-instruct":{"id":"qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"mistral-large-2411":{"id":"mistral-large-2411","name":"Mistral-Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-11-18","last_updated":"2024-11-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/mistral-large-2411"},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"sonar-reasoning-pro":{"id":"sonar-reasoning-pro","name":"Perplexity Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"sonar":{"id":"sonar","name":"Perplexity Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"claude-opus-4":{"id":"claude-opus-4","name":"Anthropic: Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-0"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.29,"output":0.59},"canonical_model_id":"alibaba/qwen3-32b"},"mistral-nemo":{"id":"mistral-nemo","name":"Mistral Nemo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16400},"cost":{"input":20,"output":40},"canonical_model_id":"mistral/mistral-nemo"},"deepseek-reasoner":{"id":"deepseek-reasoner","name":"DeepSeek Reasoner","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":0.56,"output":1.68,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-reasoner"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Google Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.3125,"cache_write":1.25},"canonical_model_id":"google/gemini-2.5-pro"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"xAI Grok Code Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":10000},"cost":{"input":0.2,"output":1.5,"cache_read":0.02},"canonical_model_id":"xai/grok-code-fast-1"},"o3-mini":{"id":"o3-mini","name":"OpenAI o3 Mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-10","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"qwen2.5-coder-7b-fast":{"id":"qwen2.5-coder-7b-fast","name":"Qwen2.5 Coder 7B fast","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-09","release_date":"2024-09-15","last_updated":"2024-09-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":0.03,"output":0.09}},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"OpenAI GPT-OSS 20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"openai/gpt-oss-20b"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"OpenAI GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"claude-4.5-haiku":{"id":"claude-4.5-haiku","name":"Anthropic: Claude 4.5 Haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"sonar-reasoning":{"id":"sonar-reasoning","name":"Perplexity Sonar Reasoning","description":"Web-grounded reasoning model for multi-step research and cited answers","family":"sonar-reasoning","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-01-27","last_updated":"2025-01-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":1,"output":5}},"claude-3.5-sonnet-v2":{"id":"claude-3.5-sonnet-v2","name":"Anthropic: Claude 3.5 Sonnet v2","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-3-5-sonnet-20241022"},"grok-3":{"id":"grok-3","name":"xAI Grok 3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.75}},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Anthropic: Claude Sonnet 4.5 (20250929)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"OpenAI GPT-4o-mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"claude-3.5-haiku":{"id":"claude-3.5-haiku","name":"Anthropic: Claude 3.5 Haiku","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":0.8,"output":4,"cache_read":0.08,"cache_write":1},"canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"llama-3.1-8b-instruct-turbo":{"id":"llama-3.1-8b-instruct-turbo","name":"Meta Llama 3.1 8B Instruct Turbo","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.02,"output":0.03},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"deepseek-v3.1-terminus":{"id":"deepseek-v3.1-terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.27,"output":1,"cache_read":0.216}},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"OpenAI GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"gpt-5-mini":{"id":"gpt-5-mini","name":"OpenAI GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"llama-prompt-guard-2-22m":{"id":"llama-prompt-guard-2-22m","name":"Meta Llama Prompt Guard 2 22M","description":"22M-parameter (DeBERTa-xsmall) English classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":2},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"meta/llama-prompt-guard-2-22m"},"gpt-5.1-chat-latest":{"id":"gpt-5.1-chat-latest","name":"OpenAI GPT-5.1 Chat","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"Anthropic: Claude Opus 4.1 (20250805)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-08","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gpt-5-pro":{"id":"gpt-5-pro","name":"OpenAI: GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":32768},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"grok-4-fast-reasoning":{"id":"grok-4-fast-reasoning","name":"xAI: Grok 4 Fast Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-reasoning"},"deepseek-v3":{"id":"deepseek-v3","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.56,"output":1.68,"cache_read":0.07}},"o1":{"id":"o1","name":"OpenAI: o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"claude-sonnet-4":{"id":"claude-sonnet-4","name":"Anthropic: Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"llama-3.1-8b-instant":{"id":"llama-3.1-8b-instant","name":"Meta Llama 3.1 8B Instant","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32678},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"hermes-2-pro-llama-3-8b":{"id":"hermes-2-pro-llama-3-8b","name":"Hermes 2 Pro Llama 3 8B","description":"Llama 3 fine-tune optimized for function calling, structured JSON responses, and dialogue","family":"hermes","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-05","release_date":"2024-04-30","last_updated":"2024-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.14,"output":0.14},"canonical_model_id":"nousresearch/hermes-2-pro-llama-3-8b"},"gpt-5.1":{"id":"gpt-5.1","name":"OpenAI GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"gpt-5-chat-latest":{"id":"gpt-5-chat-latest","name":"OpenAI GPT-5 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-09","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-chat-latest"},"grok-4":{"id":"grok-4","name":"xAI Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-09","last_updated":"2024-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":3,"output":15,"cache_read":0.75}},"claude-3-haiku-20240307":{"id":"claude-3-haiku-20240307","name":"Anthropic: Claude 3 Haiku","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2024-03-13","last_updated":"2024-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":4096},"cost":{"input":0.25,"output":1.25,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"anthropic/claude-3-haiku-20240307"},"claude-3.7-sonnet":{"id":"claude-3.7-sonnet","name":"Anthropic: Claude 3.7 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"gpt-4o":{"id":"gpt-4o","name":"OpenAI GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-05","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"xAI Grok 4.1 Fast Non-Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"llama-4-maverick":{"id":"llama-4-maverick","name":"Meta Llama 4 Maverick 17B 128E","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"o4-mini":{"id":"o4-mini","name":"OpenAI o4 Mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"sonar-deep-research":{"id":"sonar-deep-research","name":"Perplexity Sonar Deep Research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-deep-research"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Google Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.075,"cache_write":0.3},"canonical_model_id":"google/gemini-2.5-flash"},"claude-4.5-sonnet":{"id":"claude-4.5-sonnet","name":"Anthropic: Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"o1-mini":{"id":"o1-mini","name":"OpenAI: o1-mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":65536},"cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"glm-4.6":{"id":"glm-4.6","name":"Zai GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.45,"output":1.5},"canonical_model_id":"zhipuai/glm-4.6"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"output":41000},"cost":{"input":0.08,"output":0.29},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"xAI Grok 4.1 Fast Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"OpenAI GPT-OSS 120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.04,"output":0.16},"canonical_model_id":"openai/gpt-oss-120b"},"kimi-k2-0905":{"id":"kimi-k2-0905","name":"Kimi K2 (09/05)","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.5,"output":2,"cache_read":0.4},"canonical_model_id":"moonshotai/kimi-k2-0905"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"cost":{"input":0.48,"output":2},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"o3-pro":{"id":"o3-pro","name":"OpenAI o3 Pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"mistral-small":{"id":"mistral-small","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.075,"output":0.2},"canonical_model_id":"mistral/mistral-small-2506"},"gpt-4.1-mini-2025-04-14":{"id":"gpt-4.1-mini-2025-04-14","name":"OpenAI GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"gemma2-9b-it":{"id":"gemma2-9b-it","name":"Google Gemma 2","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2024-06-25","last_updated":"2024-06-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.01,"output":0.03}},"gpt-5":{"id":"gpt-5","name":"OpenAI GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"OpenAI GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"OpenAI: GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"o3":{"id":"o3","name":"OpenAI o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"deepseek-r1-distill-llama-70b":{"id":"deepseek-r1-distill-llama-70b","name":"DeepSeek R1 Distill Llama 70B","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-01-20","last_updated":"2025-01-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.03,"output":0.13}},"deepseek-tng-r1t2-chimera":{"id":"deepseek-tng-r1t2-chimera","name":"DeepSeek TNG R1T2 Chimera","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek-thinking","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-02","last_updated":"2025-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":130000,"output":163840},"cost":{"input":0.3,"output":1.2}},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"xAI Grok 4 Fast Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":16384},"cost":{"input":0.14,"output":1.4},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"chatgpt-4o-latest":{"id":"chatgpt-4o-latest","name":"OpenAI ChatGPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2024-08-14","last_updated":"2024-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":5,"output":20,"cache_read":2.5}},"claude-4.5-opus":{"id":"claude-4.5-opus","name":"Anthropic: Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.27,"output":0.41},"canonical_model_id":"deepseek/deepseek-v3.2"},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Anthropic: Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-08","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"gpt-4.1":{"id":"gpt-4.1","name":"OpenAI GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5-codex":{"id":"gpt-5-codex","name":"OpenAI: GPT-5 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"qwen3-235b-a22b-thinking":{"id":"qwen3-235b-a22b-thinking","name":"Qwen3 235B A22B Thinking","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.3,"output":2.9},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen3-coder":{"id":"qwen3-coder","name":"Qwen3 Coder 480B A35B Instruct Turbo","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.22,"output":0.95},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"Google Gemini 3 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-preview"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Google Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.025,"cache_write":0.1},"canonical_model_id":"google/gemini-2.5-flash-lite"},"kimi-k2-0711":{"id":"kimi-k2-0711","name":"Kimi K2 (07/11)","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"}}},"qihang-ai":{"id":"qihang-ai","env":["QIHANG_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.qhaigc.net/v1","name":"QiHang","doc":"https://www.qhaigc.net/docs","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0.14,"output":0.71},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.14,"output":1.14},"canonical_model_id":"openai/gpt-5.2-codex"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":0.71,"output":3.57},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":0.43,"output":2.14},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.25,"output":2},"canonical_model_id":"openai/gpt-5.2"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5-Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0.04,"output":0.29},"canonical_model_id":"openai/gpt-5-mini"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.07,"output":0.43,"tiers":[{"input":0.07,"output":0.43,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.07,"output":0.43}},"canonical_model_id":"google/gemini-3-flash-preview"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.09,"output":0.71,"tiers":[{"input":0.09,"output":0.71,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.09,"output":0.71}},"canonical_model_id":"google/gemini-2.5-flash"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"Gemini 3 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.57,"output":3.43},"canonical_model_id":"google/gemini-3-pro-preview"}}},"chutes":{"id":"chutes","env":["CHUTES_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://llm.chutes.ai/v1","name":"Chutes","doc":"https://llm.chutes.ai/v1/models","models":{"Nemotron-3-Nano-Omni-30B-TEE":{"id":"Nemotron-3-Nano-Omni-30B-TEE","name":"Nemotron 3 Nano Omni 30B TEE","description":"Omni-modal model for text, vision, audio, and multimodal agent tasks","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":0},"cost":{"input":0.0245,"output":0.0978,"cache_read":0.0024499999999999995}},"Qwen/Qwen3.6-27B-TEE":{"id":"Qwen/Qwen3.6-27B-TEE","name":"Qwen3.6 27B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.024999999999999994},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3-235B-A22B-Thinking-2507-TEE":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507-TEE","name":"Qwen3 235B A22B Thinking 2507 TEE","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2989,"output":1.1957,"cache_read":0.029889999999999993},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3.8-27B-TEE":{"id":"Qwen/Qwen3.8-27B-TEE","name":"Qwen3.8 27B TEE","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.039,"output":1.25,"cache_read":0.003899999999999999},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.5-397B-A17B-TEE":{"id":"Qwen/Qwen3.5-397B-A17B-TEE","name":"Qwen3.5 397B A17B TEE","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.45,"output":3,"cache_read":0.04499999999999999},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3-32B-TEE":{"id":"Qwen/Qwen3-32B-TEE","name":"Qwen3 32B TEE","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":40960},"cost":{"input":0.104,"output":0.416,"cache_read":0.010399999999999998},"canonical_model_id":"alibaba/qwen3-32b"},"google/gemma-4-31B-turbo-TEE":{"id":"google/gemma-4-31B-turbo-TEE","name":"gemma 4 31B turbo TEE","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.12,"output":0.37,"cache_read":0.011999999999999997},"canonical_model_id":"google/gemma-4-31b-it"},"deepseek-ai/DeepSeek-V3.2-TEE":{"id":"deepseek-ai/DeepSeek-V3.2-TEE","name":"DeepSeek V3.2 TEE","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":1,"output":1,"cache_read":0.09999999999999998},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-V4-Flash-0731-TEE":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731-TEE","name":"DeepSeek V4 Flash 0731 TEE","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32,"cache_read":0.04399999999999999},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"moonshotai/Kimi-K3-TEE":{"id":"moonshotai/Kimi-K3-TEE","name":"Kimi K3 TEE","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65535},"cost":{"input":1.85,"output":9.5,"cache_read":0.18499999999999997},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.6-TEE":{"id":"moonshotai/Kimi-K2.6-TEE","name":"Kimi K2.6 TEE","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65535},"cost":{"input":0.45,"output":1.95,"cache_read":0.04499999999999999},"canonical_model_id":"moonshotai/kimi-k2.6"},"unsloth/Mistral-Nemo-Instruct-2407-TEE":{"id":"unsloth/Mistral-Nemo-Instruct-2407-TEE","name":"Mistral Nemo Instruct 2407 TEE","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0245,"output":0.0978,"cache_read":0.0024499999999999995},"canonical_model_id":"mistral/mistral-nemo"},"zai-org/GLM-5.2-TEE":{"id":"zai-org/GLM-5.2-TEE","name":"GLM 5.2 TEE","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65535},"cost":{"input":1.25,"output":3.95,"cache_read":0.12499999999999997},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5.1-TEE":{"id":"zai-org/GLM-5.1-TEE","name":"GLM 5.1 TEE","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":65535},"cost":{"input":0.98,"output":3.08,"cache_read":0.09799999999999998},"canonical_model_id":"zhipuai/glm-5.1"}}},"gmicloud":{"id":"gmicloud","env":["GMICLOUD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.gmi-serving.com/v1","name":"GMI Cloud","doc":"https://docs.gmicloud.ai/inference-engine/api-reference/llm-api-reference","models":{"Qwen/Qwen3.8-Max":{"id":"Qwen/Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"Qwen/Qwen3.7-Max":{"id":"Qwen/Qwen3.7-Max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"Qwen/Qwen3.8-Flash":{"id":"Qwen/Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048575,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24}},"canonical_model_id":"minimax/MiniMax-M3"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.392,"output":2.784,"cache_read":0.116},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":384000},"cost":{"input":0.112,"output":0.224,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.855,"output":3.6,"cache_read":0.144},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":409600,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":409600,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":409600,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"zai-org/GLM-5.1-FP8":{"id":"zai-org/GLM-5.1-FP8","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.98,"output":3.08,"cache_read":0.182},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5.2-FP8":{"id":"zai-org/GLM-5.2-FP8","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.979,"output":3.08,"cache_read":0.182},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5-FP8":{"id":"zai-org/GLM-5-FP8","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.6,"output":1.92,"cache_read":0.12},"canonical_model_id":"zhipuai/glm-5"}}},"oci":{"id":"oci","env":["OCI_GENAI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.generativeai.us-chicago-1.oci.oraclecloud.com/openai/v1","name":"OCI Generative AI","doc":"https://docs.oracle.com/en-us/iaas/Content/generative-ai/pretrained-models.htm","models":{"meta.llama-4-scout-17b-16e-instruct":{"id":"meta.llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":16384},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"xai.grok-4.20-non-reasoning":{"id":"xai.grok-4.20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai.grok-4.3":{"id":"xai.grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"openai.gpt-oss-120b":{"id":"openai.gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"openai.gpt-oss-20b":{"id":"openai.gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"xai.grok-4.20-reasoning":{"id":"xai.grok-4.20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"meta.llama-3.3-70b-instruct":{"id":"meta.llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"xai.grok-4.6":{"id":"xai.grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"meta.llama-4-maverick-17b-128e-instruct-fp8":{"id":"meta.llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":16384},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"}}},"zenifra":{"id":"zenifra","env":["ZENIFRA_AI_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ai.zenifra.com/v1","name":"Zenifra","doc":"https://docs.zenifra.com","models":{"alibaba/qwen3.6-35b-a3b":{"id":"alibaba/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"shape":"completions"},"cost":{"input":0.19,"output":0.48},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"gitlab":{"id":"gitlab","env":["GITLAB_TOKEN"],"npm":"gitlab-ai-provider","name":"GitLab Duo","doc":"https://docs.gitlab.com/user/duo_agent_platform/","models":{"duo-chat-sonnet-5-5":{"id":"duo-chat-sonnet-5-5","name":"Agentic Chat (Claude Sonnet 5.5)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"duo-chat-opus-5":{"id":"duo-chat-opus-5","name":"Agentic Chat (Claude Opus 5)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-5"},"duo-chat-gpt-5-6-sol":{"id":"duo-chat-gpt-5-6-sol","name":"Agentic Chat (GPT-5.6 Sol)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-5.6-sol"},"duo-chat-opus-4-6":{"id":"duo-chat-opus-4-6","name":"Agentic Chat (Claude Opus 4.6)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-6"},"duo-chat-fable-5-1":{"id":"duo-chat-fable-5-1","name":"Agentic Chat (Claude Fable 5.1)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-fable-5-1"},"duo-chat-sonnet-5":{"id":"duo-chat-sonnet-5","name":"Agentic Chat (Claude Sonnet 5)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-5"},"duo-chat-gpt-5-4":{"id":"duo-chat-gpt-5-4","name":"Agentic Chat (GPT-5.4)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4"},"duo-chat-opus-5-5":{"id":"duo-chat-opus-5-5","name":"Agentic Chat (Claude Opus 5.5)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-5-5"},"duo-chat-haiku-4-5":{"id":"duo-chat-haiku-4-5","name":"Agentic Chat (Claude Haiku 4.5)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"duo-chat-gpt-6-luna":{"id":"duo-chat-gpt-6-luna","name":"Agentic Chat (GPT-6 Luna)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6-luna"},"duo-chat-gpt-5-6-luna":{"id":"duo-chat-gpt-5-6-luna","name":"Agentic Chat (GPT-5.6 Luna)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-5.6-luna"},"duo-chat-gpt-6-astra":{"id":"duo-chat-gpt-6-astra","name":"Agentic Chat (GPT-6 Astra)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6-astra"},"duo-chat-gpt-6-1-sol":{"id":"duo-chat-gpt-6-1-sol","name":"Agentic Chat (GPT-6.1 Sol)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6.1-sol"},"duo-chat-gpt-5-1":{"id":"duo-chat-gpt-5-1","name":"Agentic Chat (GPT-5.1)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.1"},"duo-chat-gpt-5-4-nano":{"id":"duo-chat-gpt-5-4-nano","name":"Agentic Chat (GPT-5.4 Nano)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4-nano"},"duo-chat-gpt-5-mini":{"id":"duo-chat-gpt-5-mini","name":"Agentic Chat (GPT-5 Mini)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5-mini"},"duo-chat-gpt-5-5":{"id":"duo-chat-gpt-5-5","name":"Agentic Chat (GPT-5.5)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.5"},"duo-chat-sonnet-4-5":{"id":"duo-chat-sonnet-4-5","name":"Agentic Chat (Claude Sonnet 4.5)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"duo-chat-gpt-5-codex":{"id":"duo-chat-gpt-5-codex","name":"Agentic Chat (GPT-5 Codex)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5-codex"},"duo-chat-opus-4-7":{"id":"duo-chat-opus-4-7","name":"Agentic Chat (Claude Opus 4.7)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-7"},"duo-chat-fable-5":{"id":"duo-chat-fable-5","name":"Agentic Chat (Claude Fable 5)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-fable-5"},"duo-chat-gpt-5-2-codex":{"id":"duo-chat-gpt-5-2-codex","name":"Agentic Chat (GPT-5.2 Codex)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.2-codex"},"duo-chat-gpt-5-2":{"id":"duo-chat-gpt-5-2","name":"Agentic Chat (GPT-5.2)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.2"},"duo-chat-sonnet-4-6":{"id":"duo-chat-sonnet-4-6","name":"Agentic Chat (Claude Sonnet 4.6)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"duo-chat-opus-4-5":{"id":"duo-chat-opus-4-5","name":"Agentic Chat (Claude Opus 4.5)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"duo-chat-gpt-6-sol":{"id":"duo-chat-gpt-6-sol","name":"Agentic Chat (GPT-6 Sol)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6-sol"},"duo-chat-gpt-5-6-terra":{"id":"duo-chat-gpt-5-6-terra","name":"Agentic Chat (GPT-5.6 Terra)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-5.6-terra"},"duo-chat-gpt-5-3-codex":{"id":"duo-chat-gpt-5-3-codex","name":"Agentic Chat (GPT-5.3 Codex)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.3-codex"},"duo-chat-gpt-5-4-mini":{"id":"duo-chat-gpt-5-4-mini","name":"Agentic Chat (GPT-5.4 Mini)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4-mini"},"duo-chat-opus-4-8":{"id":"duo-chat-opus-4-8","name":"Agentic Chat (Claude Opus 4.8)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-8"}}},"bailing":{"id":"bailing","env":["BAILING_API_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tbox.cn/api/llm/v1","name":"Bailing","doc":"https://alipaytbox.yuque.com/sxs0ba/ling/intro","models":{"Ring-1T":{"id":"Ring-1T","name":"Ring-1T","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"ring","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-10","last_updated":"2025-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.57,"output":2.29}},"Ling-1T":{"id":"Ling-1T","name":"Ling-1T","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-10","last_updated":"2025-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.57,"output":2.29}}}},"hyper":{"id":"hyper","env":["HYPER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://hyper.charm.land/v1","name":"Charm Hyper","doc":"https://hyper.charm.land","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0.33,"output":1.31,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.52432,"output":4.79072,"cache_read":0.283088},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16000},"cost":{"input":3.2664,"output":16.332,"cache_read":0.32664},"canonical_model_id":"moonshotai/kimi-k3"},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":64000},"cost":{"input":0.2,"output":0.8,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.7-max"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16000},"cost":{"input":1.03436,"output":4.3552,"cache_read":0.206872},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":6553},"cost":{"input":0.484,"output":1.852,"cache_read":0.242},"canonical_model_id":"minimax/MiniMax-M2.7"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":25600},"cost":{"input":0.098,"output":0.334,"cache_read":0.049},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"prism":{"id":"prism","name":"Prism","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"release_date":"1970-01-01","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":16000},"cost":{"input":3.2664,"output":16.332,"cache_read":0.32664}},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0.33,"output":1.31,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":1.437216,"output":4.311648,"cache_read":0.047907},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"cost":{"input":0.32664,"output":1.30656,"cache_read":0.064239},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":1.52432,"output":4.79072,"cache_read":0.152432},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.16332,"output":0.5444,"cache_read":0.031575},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.2,"output":4.8,"cache_read":0.24},"canonical_model_id":"alibaba/qwen3.7-plus"},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":1.0888,"output":4.40964,"cache_read":0.185096},"canonical_model_id":"thinkingmachines/inkling"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":13107},"cost":{"input":0.178,"output":0.68,"cache_read":0.089},"canonical_model_id":"openai/gpt-oss-120b"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":26214},"cost":{"input":0.6,"output":2.5,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k2-thinking"}}},"tencent-coding-plan":{"id":"tencent-coding-plan","env":["TENCENT_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lkeap.cloud.tencent.com/coding/v3","name":"Tencent Coding Plan (China)","doc":"https://cloud.tencent.com/document/product/1772/128947","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"hunyuan-2.0-thinking":{"id":"hunyuan-2.0-thinking","name":"Tencent HY 2.0 Think","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M2.5"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"hunyuan-2.0-instruct":{"id":"hunyuan-2.0-instruct","name":"Tencent HY 2.0 Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"hunyuan-t1":{"id":"hunyuan-t1","name":"Hunyuan-T1","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"tc-code-latest":{"id":"tc-code-latest","name":"Auto","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"hunyuan-turbos":{"id":"hunyuan-turbos","name":"Hunyuan-TurboS","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}}}},"aihubmix":{"id":"aihubmix","env":["AIHUBMIX_API_KEY"],"npm":"@aihubmix/ai-sdk-provider","name":"AIHubMix","doc":"https://docs.aihubmix.com","models":{"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.155,"output":0.62,"cache_read":0.0031},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"xiaomi-mimo-v2.5":{"id":"xiaomi-mimo-v2.5","name":"Xiaomi MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.44,"output":2.2,"cache_read":0.088,"tiers":[{"input":0.88,"output":4.4,"cache_read":0.176,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.88,"output":4.4,"cache_read":0.176}},"canonical_model_id":"xiaomi/mimo-v2.5"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1268,"output":3.9438,"cache_read":0.2817},"canonical_model_id":"zhipuai/glm-5.3"},"coding-minimax-m2.7":{"id":"coding-minimax-m2.7","name":"Coding MiniMax M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128100},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"coding-glm-5.3-free":{"id":"coding-glm-5.3-free","name":"GLM-5.3 (Coding, free)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"doubao-seed-2-0-mini-260428":{"id":"doubao-seed-2-0-mini-260428","name":"Doubao Seed 2.0 Mini 260428","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.030136,"output":0.30136,"cache_read":0.0060272,"input_audio":0.423,"tiers":[{"input":0.060274,"output":0.60274,"cache_read":0.0120548,"tier":{"type":"context","size":32000}},{"input":0.120548,"output":1.20548,"cache_read":0.0241096,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"doubao-seed-2-0-pro":{"id":"doubao-seed-2-0-pro","name":"Doubao Seed 2.0 Pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4958,"output":2.479,"cache_read":0.09916,"tiers":[{"input":0.7436,"output":3.718,"cache_read":0.14872,"tier":{"type":"context","size":32000}},{"input":1.4872,"output":7.436,"cache_read":0.29744,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"ling-3.0-flash-free":{"id":"ling-3.0-flash-free","name":"Ling 3.0 Flash (free)","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"inclusionai/ling-3.0-flash"},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.375,"output":4.675},"canonical_model_id":"meta/muse-spark-1.1"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":65536},"cost":{"input":0.282,"output":1.692,"cache_read":0.0282,"cache_write":0.3525,"tiers":[{"input":1.128,"output":6.768,"cache_read":0.1128,"cache_write":1.41,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.128,"output":6.768,"cache_read":0.1128,"cache_write":1.41}},"canonical_model_id":"alibaba/qwen3.6-plus"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"gpt-image-2.5-flare":{"id":"gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"qwen3.8-max-preview":{"id":"qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.338,"output":1.014,"cache_read":0.0676,"cache_write":0.4225},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"nemotron-3.5-lightning-free":{"id":"nemotron-3.5-lightning-free","name":"Nemotron 3.5 Lightning 30B A3B (free)","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"input":991808,"output":131072},"cost":{"input":0.0282,"output":0.1128,"cache_read":0.00564,"cache_write":0.03525,"tiers":[{"input":0.0845,"output":0.338,"cache_read":0.0169,"cache_write":0.105625,"tier":{"type":"context","size":32000}},{"input":0.169,"output":0.676,"cache_read":0.0338,"cache_write":0.21125,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.375,"output":4.675},"canonical_model_id":"meta/muse-spark-1.2"},"deepseek-v4-flash-0731-fast":{"id":"deepseek-v4-flash-0731-fast","name":"DeepSeek V4 Flash 0731 Fast","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.28,"output":1.4,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.274,"output":1.096,"tiers":[{"input":0.548,"output":1.644,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5v"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"gemini-3.1-pro-preview-customtools":{"id":"gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":131072},"cost":{"input":1.69,"output":5.07,"cache_read":0.169,"cache_write":2.1125},"canonical_model_id":"alibaba/qwen3.7-max"},"gpt-image-2.5-sunburst":{"id":"gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.155,"output":0.31,"cache_read":0.0031},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"doubao-seed-2-0-code-preview":{"id":"doubao-seed-2-0-code-preview","name":"Doubao Seed 2.0 Code Preview","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4822,"output":2.411,"cache_read":0.09644,"tiers":[{"input":0.72328,"output":3.6164,"cache_read":0.144656,"tier":{"type":"context","size":32000}},{"input":1.4466,"output":7.233,"cache_read":0.28932,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"alicloud-glm-5.1":{"id":"alicloud-glm-5.1","name":"GLM-5.1 (Alibaba Cloud)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":0.84,"output":3.38,"cache_read":0.169,"cache_write":1.05625},"canonical_model_id":"zhipuai/glm-5.1"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.1562,"output":0.6248,"cache_read":0.03905},"canonical_model_id":"tencent/hy3"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"nemotron-3-nano-omni-30b-a3b-reasoning-free":{"id":"nemotron-3-nano-omni-30b-a3b-reasoning-free","name":"Nemotron 3 Nano Omni 30B A3B Reasoning (free)","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.0564,"output":0.4512,"tiers":[{"input":0.2254,"output":1.8032,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.137,"output":1.37,"cache_read":0,"tiers":[{"input":0.2054,"output":2.054,"tier":{"type":"context","size":32000}},{"input":0.411,"output":4.11,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-vl-plus"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"claude-opus-4-8-think":{"id":"claude-opus-4-8-think","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-opus-4-7-think":{"id":"claude-opus-4-7-think","name":"Claude Opus 4.7 Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"nemotron-3-ultra-550b-a55b-free":{"id":"nemotron-3-ultra-550b-a55b-free","name":"Nemotron 3 Ultra 550B A55B (free)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"qwen3.8-max-free":{"id":"qwen3.8-max-free","name":"Qwen3.8 Max (free)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.8-max"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":3.9995,"cache_read":0.160835},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"coding-glm-5.1-free":{"id":"coding-glm-5.1-free","name":"Coding GLM 5.1 (free)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.1"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"hy3-preview":{"id":"hy3-preview","name":"Hy3 Preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.17,"output":0.566661,"cache_read":0.051},"canonical_model_id":"tencent/hy3-preview"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.499999,"cache_read":0.03,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.22,"output":1.32,"cache_read":0.044},"canonical_model_id":"stepfun/step-3.7-flash"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.4508,"output":2.7048,"cache_read":0.09016,"cache_write":0.5635,"tiers":[{"input":0.902,"output":5.412,"cache_read":0.1804,"cache_write":1.1275,"tier":{"type":"context","size":32000}},{"input":1.3522,"output":8.1132,"cache_read":0.27044,"cache_write":1.69025,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"claude-sonnet-4-5-think":{"id":"claude-sonnet-4-5-think","name":"Claude Sonnet 4.5 Thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"alicloud-deepseek-v4-pro":{"id":"alicloud-deepseek-v4-pro","name":"DeepSeek V4 Pro (Alibaba Cloud)","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":1.69,"output":3.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen3.8-omni-flash":{"id":"qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1126,"output":0.380025,"cache_read":0.014075,"cache_write":0.1759375},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.48,"output":0.96,"cache_read":0.00384},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.273974,"output":1.095896,"cache_read":0.0547948,"tiers":[{"input":0.547946,"output":2.191784,"cache_read":0.1095892,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"qwen3.5-122b-a10b":{"id":"qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1126,"output":0.9008,"tiers":[{"input":0.2818,"output":2.2544,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"coding-glm-5.3-flash-free":{"id":"coding-glm-5.3-flash-free","name":"GLM-5.3-Flash (Coding, free)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"coding-xiaomi-mimo-v2.5-pro":{"id":"coding-xiaomi-mimo-v2.5-pro","name":"Coding Xiaomi MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.2,"output":0.4,"cache_read":0.0016},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.6918,"output":2.0754,"cache_read":0.023057694},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":0.275,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5-1"},"zai-glm-5.1":{"id":"zai-glm-5.1","name":"GLM-5.1 (Z.ai)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":0.845,"output":3.38,"cache_read":0.183112},"canonical_model_id":"zhipuai/glm-5.1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.155,"output":0.62,"cache_read":0.0031},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"qwen3.8-flash-free":{"id":"qwen3.8-flash-free","name":"Qwen3.8 Flash (free)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"longcat-2.0":{"id":"longcat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.7746,"output":3.0984,"cache_read":0.015492},"canonical_model_id":"meituan/longcat-2.0"},"qwen3.6-max-preview":{"id":"qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":240000,"output":64000},"status":"deprecated","cost":{"input":1.27,"output":7.61,"cache_read":0.1268,"cache_write":1.585,"tiers":[{"input":2.11,"output":12.67,"cache_read":0.2112,"cache_write":2.64,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.375,"output":4.675,"cache_read":0.165},"canonical_model_id":"meta/muse-spark-1.3"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"glm-4.7-flash-free":{"id":"glm-4.7-flash-free","name":"GLM-4.7-Flash (free)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.155,"output":0.31,"cache_read":0.0031},"canonical_model_id":"xiaomi/mimo-v2.5"},"coding-glm-5.2-free":{"id":"coding-glm-5.2-free","name":"GLM-5.2 (Coding, free)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1.08,"output":3.0888,"cache_read":0.054},"canonical_model_id":"stepfun/step-5-preview"},"coding-xiaomi-mimo-v2.5":{"id":"coding-xiaomi-mimo-v2.5","name":"Coding Xiaomi MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.08,"output":0.16,"cache_read":0.0016},"canonical_model_id":"xiaomi/mimo-v2.5"},"nemotron-3-super-120b-a12b-free":{"id":"nemotron-3-super-120b-a12b-free","name":"Nemotron 3 Super 120B A12B (free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"deep-deepseek-v4-flash":{"id":"deep-deepseek-v4-flash","name":"DeepSeek V4 Flash (DeepSeek)","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.154,"output":0.308,"cache_read":0.0308},"canonical_model_id":"deepseek/deepseek-v4-flash"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1644,"output":0.9864,"tiers":[{"input":0.411,"output":2.466,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.142,"output":0.284,"cache_read":0.0284},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"mimo-v2.6-pro-ultraspeed":{"id":"mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.8,"output":9.6,"cache_read":0.0384},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":131072},"cost":{"input":1.69,"output":5.07,"cache_read":0.169,"cache_write":2.1125},"canonical_model_id":"alibaba/qwen3.8-max"},"xiaomi-mimo-v2.5-free":{"id":"xiaomi-mimo-v2.5-free","name":"Xiaomi MiMo-V2.5 (free)","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":512000},"cost":{"input":0.288,"output":1.152},"canonical_model_id":"minimax/MiniMax-M3"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1096,"output":0.6576,"cache_read":0.01096,"cache_write":0.137,"tiers":[{"input":0.274,"output":1.644,"cache_read":0.0274,"cache_write":0.3425,"tier":{"type":"context","size":128000}},{"input":0.548,"output":3.288,"cache_read":0.0548,"cache_write":0.685,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.5-plus"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.48,"output":0.96,"cache_read":0.00384},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1268,"output":3.9438,"cache_read":0.2817},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"alicloud-deepseek-v4-flash":{"id":"alicloud-deepseek-v4-flash","name":"DeepSeek V4 Flash (Alibaba Cloud)","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1126,"output":0.380025,"cache_read":0.014075,"cache_write":0.1759375},"canonical_model_id":"alibaba/qwen3.8-flash"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11268,"output":0.39438,"cache_read":0.02817},"canonical_model_id":"zhipuai/glm-5.3-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":64000},"cost":{"input":0.282,"output":1.128,"cache_read":0.0564,"cache_write":0.3525,"tiers":[{"input":0.845,"output":3.38,"cache_read":0.169,"cache_write":1.05625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.845,"output":3.38,"cache_read":0.169,"cache_write":1.05625}},"canonical_model_id":"alibaba/qwen3.7-plus"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"interleaved":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image"},"claude-sonnet-4-6-think":{"id":"claude-sonnet-4-6-think","name":"Claude Sonnet 4.6 Thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"nemotron-3.5-content-safety-free":{"id":"nemotron-3.5-content-safety-free","name":"Nemotron 3.5 Content Safety (free)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"claude-opus-4-5-think":{"id":"claude-opus-4-5-think","name":"Claude Opus 4.5 Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.499,"cache_read":0.03,"input_audio":0.9999},"canonical_model_id":"google/gemini-2.5-flash"},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"interleaved":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM 5 Vision Turbo","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.7042,"output":3.09848,"cache_read":0.169008,"tiers":[{"input":0.986,"output":3.662004,"cache_read":0.253402,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"coding-minimax-m2.7-highspeed":{"id":"coding-minimax-m2.7-highspeed","name":"Coding MiniMax M2.7 Highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128100},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.273974,"output":1.095896,"cache_read":0.0547948,"tiers":[{"input":0.547946,"output":2.191784,"cache_read":0.1095892,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.137,"output":0.411,"cache_read":0.0274,"tiers":[{"input":0.274,"output":0.822,"cache_read":0.0548,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.845,"output":2.535,"cache_read":0.04225},"canonical_model_id":"tencent/hy4-preview"},"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"GPT-Image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":32},"canonical_model_id":"openai/gpt-image-1.5"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"qwen3.5-flash":{"id":"qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.0282,"output":0.282,"cache_read":0.00282,"cache_write":0.03525,"tiers":[{"input":0.1126,"output":1.126,"cache_read":0.01126,"cache_write":0.14075,"tier":{"type":"context","size":128000}},{"input":0.169,"output":1.69,"cache_read":0.0169,"cache_write":0.21125,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.5-flash"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"claude-opus-4-6-think":{"id":"claude-opus-4-6-think","name":"Claude Opus 4.6 Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"coding-glm-5.1":{"id":"coding-glm-5.1","name":"Coding GLM 5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.06,"output":0.22,"cache_read":0.013},"canonical_model_id":"zhipuai/glm-5.1"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.0846,"output":0.6768,"tiers":[{"input":0.2536,"output":2.0288,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"xiaomi-mimo-v2.5-pro":{"id":"xiaomi-mimo-v2.5-pro","name":"Xiaomi MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.1,"output":3.3,"cache_read":0.22,"tiers":[{"input":2.2,"output":6.6,"cache_read":0.44,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2.2,"output":6.6,"cache_read":0.44}},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"gemini-3.1-flash-lite-image":{"id":"gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5,"cache_read":0.025},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"coding-minimax-m2.7-free":{"id":"coding-minimax-m2.7-free","name":"Coding MiniMax M2.7 (Free)","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128100},"cost":{"input":0,"output":0},"canonical_model_id":"minimax/MiniMax-M2.7"},"deep-deepseek-v4-pro":{"id":"deep-deepseek-v4-pro","name":"DeepSeek V4 Pro (DeepSeek)","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.478,"output":0.956,"cache_read":0.004302},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.5-flash"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.254,"output":1.524},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"xiaomi-mimo-v2.5-pro-free":{"id":"xiaomi-mimo-v2.5-pro-free","name":"Xiaomi MiMo-V2.5-Pro (free)","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":1000000},"cost":{"input":2.2,"output":6.6,"cache_read":0.55,"tiers":[{"input":4.4,"output":13.2,"cache_read":1.1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.4,"output":13.2,"cache_read":1.1}},"canonical_model_id":"xai/grok-4.5"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash-image"},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":65536},"cost":{"input":0.169,"output":1.014,"cache_read":0.0169,"cache_write":0.21125,"tiers":[{"input":0.676,"output":4.056,"cache_read":0.0676,"cache_write":0.845,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.676,"output":4.056,"cache_read":0.0676,"cache_write":0.845}},"canonical_model_id":"alibaba/qwen3.6-flash"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.9,"output":7.999,"cache_read":0.32167},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"doubao-seed-2-0-lite-260428":{"id":"doubao-seed-2-0-lite-260428","name":"Doubao Seed 2.0 Lite 260428","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.09041,"output":0.54246,"cache_read":0.018082,"input_audio":1.35615,"tiers":[{"input":0.135616,"output":0.813696,"cache_read":0.0271232,"tier":{"type":"context","size":32000}},{"input":0.271232,"output":1.627392,"cache_read":0.0542464,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"}}},"commandcode":{"id":"commandcode","env":["CMD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.commandcode.ai/provider/v1","name":"CommandCode","doc":"https://commandcode.ai/docs/provider","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":0},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2,"output":8,"cache_read":0.5,"cache_write":0},"canonical_model_id":"openai/gpt-5.3-codex"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":0},"canonical_model_id":"openai/gpt-5.5"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0},"canonical_model_id":"openai/gpt-5.4-mini"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4.1-flash-fast":{"id":"deepseek/deepseek-v4.1-flash-fast","name":"DeepSeek V4.1 Flash Fast","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.16,"output":0.58,"cache_read":0.016},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-v4-flash-fast":{"id":"deepseek/deepseek-v4-flash-fast","name":"DeepSeek V4 Flash Fast","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"Qwen/Qwen3.8-Max":{"id":"Qwen/Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"Qwen/Qwen3.8-Omni-Flash":{"id":"Qwen/Qwen3.8-Omni-Flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"Qwen/Qwen3.8-Max-0902":{"id":"Qwen/Qwen3.8-Max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"Qwen/Qwen3.8-Flash":{"id":"Qwen/Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":3,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":0.5,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"inclusionai/ling-3.1-flash:free":{"id":"inclusionai/ling-3.1-flash:free","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.1-flash"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"cache_write":0.08334},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"stepfun/Step-3.5-Flash":{"id":"stepfun/Step-3.5-Flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.09,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"stepfun/Step-3.7-Flash":{"id":"stepfun/Step-3.7-Flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/Step-5-Preview":{"id":"stepfun/Step-5-Preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"mistral/mistral-large-4":{"id":"mistral/mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.7-Code-Highspeed":{"id":"moonshotai/Kimi-K2.7-Code-Highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 (Contributor)","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 (Contributor)","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy3-paid":{"id":"tencent/hy3-paid","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.2-Fast":{"id":"zai-org/GLM-5.2-Fast","name":"GLM-5.2-Fast","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":10.25,"cache_read":0.5},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"orcarouter":{"id":"orcarouter","env":["ORCAROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.orcarouter.ai/v1","name":"OrcaRouter","doc":"https://docs.orcarouter.ai","models":{"grok/grok-4.3":{"id":"grok/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"grok/grok-4.6":{"id":"grok/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"grok/grok-4.5":{"id":"grok/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-reasoner":{"id":"deepseek/deepseek-reasoner","name":"DeepSeek Reasoner","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.242,"output":0.726,"cache_read":0.0077},"canonical_model_id":"deepseek/deepseek-reasoner"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash-free":{"id":"deepseek/deepseek-v4-flash-free","name":"DeepSeek V4 Flash (free)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-chat"},"kimi/kimi-k3":{"id":"kimi/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"kimi/kimi-k2.5":{"id":"kimi/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3,"cache_read":0.1,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"kimi/kimi-k2.7-code":{"id":"kimi/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi/kimi-k2.6":{"id":"kimi/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"google/gemini-flash-latest"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":18,"cache_read":0.4},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.33,"cache_read":0.0075},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-robotics-er-1.6-preview":{"id":"google/gemini-robotics-er-1.6-preview","name":"Gemini Robotics-ER 1.6 Preview","description":"Vision-language model for embodied reasoning: spatial understanding, task planning, and physical-world agentic robotics","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-14","last_updated":"2026-04-14","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":1,"output":5},"canonical_model_id":"google/gemini-robotics-er-1.6-preview"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.13,"output":0.38,"cache_read":0.02},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.08333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-flash-lite-latest"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5,"cache_read":0},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-5.2-chat-latest":{"id":"openai/gpt-5.2-chat-latest","name":"GPT-5.2 Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":100000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.1-chat-latest":{"id":"openai/gpt-5.1-chat-latest","name":"GPT-5.1 Chat","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5-chat-latest":{"id":"openai/gpt-5-chat-latest","name":"GPT-5 Chat (latest)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":100000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-chat-latest"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.03,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.33,"output":2.4},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3-vl-235b-a22b-thinking":{"id":"qwen/qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.4,"output":4},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.276,"output":1.651,"cache_read":0.0276,"cache_write":0.345,"tiers":[{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":131072},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":3.75,"cache_read":0.25,"cache_write":1.563},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.057,"output":0.459,"tiers":[{"input":0.229,"output":1.835,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.359,"output":1.434,"tiers":[{"input":0.574,"output":2.294,"tier":{"type":"context","size":32000}},{"input":1.004,"output":4.014,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.115,"output":0.917,"tiers":[{"input":0.287,"output":2.294,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.172,"output":1.032,"tiers":[{"input":0.43,"output":2.58,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"tiers":[{"input":0.5,"output":3,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.5,"output":3}},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.35,"output":1.42,"cache_read":0.071,"tiers":[{"input":1.06,"output":4.25,"cache_read":0.21,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.06,"output":4.25,"cache_read":0.21}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.086,"output":0.688,"tiers":[{"input":0.258,"output":2.064,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"tiers":[{"input":1,"output":4,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1,"output":4}},"canonical_model_id":"alibaba/qwen3.6-flash"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.18,"output":0.59,"cache_read":0.059},"canonical_model_id":"tencent/hy3"},"tencent/hy3-free":{"id":"tencent/hy3-free","name":"Hy3 (free)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"tencent/hy3"},"orcarouter/auto":{"id":"orcarouter/auto","name":"OrcaRouter Auto","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2026-05-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0,"output":0}},"orcarouter/fusion-flash":{"id":"orcarouter/fusion-flash","name":"OrcaRouter Fusion Flash","description":"Budget Fusion panel that runs Gemini 3.5 Flash, MiniMax M2.7 and GLM 5.1 in parallel on hard requests, then has a Claude Opus 4.8 judge return the strongest single answer verbatim. Cost-sensitive fan-out over a 200K window.","family":"model-router","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000}},"orcarouter/fusion":{"id":"orcarouter/fusion","name":"OrcaRouter Fusion","description":"Curated fan-out router that runs Claude Opus 4.8, GPT-5.5 and Gemini 3.1 Pro in parallel on hard requests, then has a Claude Opus 4.8 judge return the strongest single answer verbatim. Easy requests fall through to a cheaper default and bill as one call.","family":"model-router","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"orcarouter/free":{"id":"orcarouter/free","name":"OrcaRouter Free","description":"Built-in router over the free tier that scores each request's difficulty and sends light work to the smaller free model and harder work to the stronger one. Priced at zero and never falls back to a paid model.","family":"auto","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":0,"output":0}},"orcarouter/fusion-mini":{"id":"orcarouter/fusion-mini","name":"OrcaRouter Fusion Mini","description":"Leaner two-model Fusion panel that runs Claude Opus 4.8 and GPT-5.5 in parallel on hard requests, then has a Claude Opus 4.8 judge return the strongest single answer verbatim. Easy requests fall through to a cheaper default and bill as one call.","family":"model-router","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":10},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.26,"output":3.96,"cache_read":0.234},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5.3-flash-free":{"id":"z-ai/glm-5.3-flash-free","name":"GLM-5.3-Flash (free)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"}}},"togetherai":{"id":"togetherai","env":["TOGETHER_API_KEY"],"npm":"@ai-sdk/togetherai","name":"Together AI","doc":"https://docs.together.ai/docs/serverless-models","models":{"Qwen/Qwen3-235B-A22B-Instruct-2507-tput":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507-tput","name":"Qwen3 235B A22B Instruct 2507 FP8","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.2,"output":0.6},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3.7-Max":{"id":"Qwen/Qwen3.7-Max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.7-max"},"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":2,"output":2},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.6-Plus":{"id":"Qwen/Qwen3.6-Plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"alibaba/qwen3.6-plus"},"Qwen/Qwen2.5-7B-Instruct-Turbo":{"id":"Qwen/Qwen2.5-7B-Instruct-Turbo","name":"Qwen 2.5 7B Instruct Turbo","description":"Small open Qwen2.5 instruction model for chat, structured output, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.3,"output":0.3},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"Qwen/Qwen3-Coder-Next-FP8":{"id":"Qwen/Qwen3-Coder-Next-FP8","name":"Qwen3 Coder Next FP8","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":130000},"status":"deprecated","cost":{"input":0.6,"output":3.6,"cache_read":0.35},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":524288},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.5"},"reflection/Beam-1.0-501B-A23B":{"id":"reflection/Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":1.2,"cache_read":0.04},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"LiquidAI/LFM2-24B-A2B":{"id":"LiquidAI/LFM2-24B-A2B","name":"LFM2-24B-A2B","description":"Efficient open Liquid MoE instruct model for on-device chat and tool calling","family":"liquid","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.03,"output":0.12},"canonical_model_id":"liquid/lfm2-24b-a2b"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":131072},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"essentialai/Rnj-1-Instruct":{"id":"essentialai/Rnj-1-Instruct","name":"Rnj-1 Instruct","description":"Open 8B Essential AI model for code, STEM reasoning, and agentic tool use","family":"rnj","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-04","last_updated":"2025-12-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.15},"canonical_model_id":"essentialai/rnj-1-instruct"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V3-1":{"id":"deepseek-ai/DeepSeek-V3-1","name":"DeepSeek V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":1.7},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":1.25,"output":1.25},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163839,"output":163839},"status":"deprecated","cost":{"input":3,"output":7},"canonical_model_id":"deepseek/deepseek-r1"},"meta-models/Muse-Spark-Open-816B-A42B":{"id":"meta-models/Muse-Spark-Open-816B-A42B","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"meta-llama/Llama-3.3-70B-Instruct-Turbo":{"id":"meta-llama/Llama-3.3-70B-Instruct-Turbo","name":"Llama 3.3 70B","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":1.04,"output":1.04},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512300,"output":512300},"status":"deprecated","cost":{"input":0.6,"output":3.6,"cache_read":0.2},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":262144},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":400000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":164000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"llmtr":{"id":"llmtr","env":["LLMTR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://llmtr.com/v1","name":"LLMTR","doc":"https://llmtr.com/docs","models":{"medgemma-4b":{"id":"medgemma-4b","name":"MedGemma 4B","description":"Multimodal medical-domain Gemma variant for text and image analysis","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-26","last_updated":"2026-08-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"status":"deprecated","cost":{"input":3,"output":5}},"magibu-11b-v8":{"id":"magibu-11b-v8","name":"Magibu 11B v8","description":"Turkish-language chat model for instruction following and assistant flows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-06-05","last_updated":"2026-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.1,"output":0.5}},"gemma-4":{"id":"gemma-4","name":"Gemma 4","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":2,"output":5,"cache_read":0.5},"canonical_model_id":"google/gemma-4-31b-it"},"muse-glimmer-30b-tr":{"id":"muse-glimmer-30b-tr","name":"Muse Glimmer 30B (TR)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":2,"output":5,"cache_read":0.5},"canonical_model_id":"meta/muse-glimmer-30b"},"trendyol-asure-12b":{"id":"trendyol-asure-12b","name":"Trendyol Asure 12B","description":"Turkish-language multimodal instruct model built on Gemma 3 12B for e-commerce text, chat, and image-text tasks","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-19","last_updated":"2026-02-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":40960},"cost":{"input":0.1,"output":0.5,"cache_read":0.025},"canonical_model_id":"trendyol/asure-12b"},"qwen3-6-35b":{"id":"qwen3-6-35b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":5,"output":10},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.58,"output":1.44,"cache_read":0.116},"canonical_model_id":"thinkingmachines/inkling-small"},"mimo/mimo-v2.5":{"id":"mimo/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.14,"output":0.28},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo/mimo-v2.5-pro":{"id":"mimo/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.435,"output":0.87},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"upstage/solar-pro3":{"id":"upstage/solar-pro3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.15,"output":0.6,"cache_read":0.03},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-pro2":{"id":"upstage/solar-pro2","name":"Solar Pro 2","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro2"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.1},"canonical_model_id":"google/gemini-2.5-flash-lite"},"mistral/voxtral-small-latest":{"id":"mistral/voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-latest"},"publicai/apertus-8b-instruct":{"id":"publicai/apertus-8b-instruct","name":"Apertus 8B Instruct","description":"Fully open 8B multilingual LLM supporting 1800+ languages with 65K context. Trained on compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"swiss-ai/apertus-8b"},"publicai/apertus-70b-instruct":{"id":"publicai/apertus-70b-instruct","name":"Apertus 70B Instruct","description":"Fully open 70B multilingual LLM supporting 1800+ languages with 65K context. Trained on 15T tokens of compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"cost":{"input":0.82,"output":2.92},"canonical_model_id":"swiss-ai/apertus-70b"},"perplexity/sonar-deep-research":{"id":"perplexity/sonar-deep-research","name":"Sonar Deep Research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3-vl-plus":{"id":"qwen/qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"cost":{"input":0.2,"output":1.6,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3-vl-plus"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":1,"output":5},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"status":"deprecated","cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen-flash":{"id":"qwen/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.05,"output":0.4,"cache_read":0.01},"canonical_model_id":"alibaba/qwen-flash"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.2,"cache_read":0.08},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"alibaba/qwen3.6-flash"}}},"opper":{"id":"opper","env":["OPPER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.opper.ai/v3/compat","name":"Opper","doc":"https://opper.ai/models","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.5811,"output":3},"canonical_model_id":"alibaba/qwen3.8-27b"},"sonar-pro":{"id":"sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.75,"output":4.6488,"cache_read":0.44},"canonical_model_id":"zhipuai/glm-5.3"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15},"canonical_model_id":"moonshotai/kimi-k3"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":1.78812,"output":3.57624},"canonical_model_id":"deepseek/deepseek-v4-pro"},"sonar-reasoning-pro":{"id":"sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"sonar":{"id":"sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.11622,"output":0.488124},"canonical_model_id":"openai/gpt-oss-20b"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.69732,"output":2.78928},"canonical_model_id":"minimax/MiniMax-M2.7"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.25,"output":0.66},"canonical_model_id":"deepseek/deepseek-v4-flash"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.12,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"tier":{"type":"context","size":524288}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24}},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":6,"cache_read":0.63},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.62708,"output":5.811},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":0.5,"cache_read":0.07},"canonical_model_id":"zhipuai/glm-5.3-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":1.1622,"output":4.88124},"canonical_model_id":"openai/gpt-oss-120b"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.46488,"output":2.44062},"canonical_model_id":"google/gemma-4-31b-it"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.5811,"output":2.44062},"canonical_model_id":"mistral/mistral-small-2603"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.825,"output":4.125,"cache_read":0.0825},"canonical_model_id":"google/gemini-3.8-flash"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.5811,"output":2.3244},"canonical_model_id":"alibaba/qwen3-coder-next"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"perplexity-agent":{"id":"perplexity-agent","env":["PERPLEXITY_API_KEY"],"npm":"@ai-sdk/openai","api":"https://api.perplexity.ai/v1","name":"Perplexity Agent","doc":"https://docs.perplexity.ai/docs/agent-api/models","models":{"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4-1-fast-non-reasoning":{"id":"xai/grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"status":"deprecated","cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"tiers":[{"input":0.5,"output":3,"cache_read":0.05,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.5,"output":3,"cache_read":0.05}},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4}},"canonical_model_id":"openai/gpt-6-sol"},"perplexity/nemotron-3.5-super-120b-a12b":{"id":"perplexity/nemotron-3.5-super-120b-a12b","name":"Nemotron 3.5 Super 120B A12B","description":"NVIDIA Nemotron 3.5 Super reasoning model (120B total, 12B active) served with text-only input","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.2,"output":0.5,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-super-120b-a12b"},"perplexity/glm-5.3":{"id":"perplexity/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"perplexity/kimi-k3":{"id":"perplexity/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.25,"output":2.5,"cache_read":0.0625},"canonical_model_id":"perplexity/sonar"},"perplexity/deepseek-v4-pro-0813":{"id":"perplexity/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"perplexity/glm-5.3-flash":{"id":"perplexity/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"moonshot-ai/kimi-k3":{"id":"moonshot-ai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot-ai/kimi-k2.7-code":{"id":"moonshot-ai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32000},"status":"deprecated","cost":{"input":0.25,"output":2.5},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"}}},"anthropic":{"id":"anthropic","env":["ANTHROPIC_API_KEY"],"npm":"@ai-sdk/anthropic","name":"Anthropic","doc":"https://docs.anthropic.com/en/docs/about-claude/models","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25}},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.4,"cache_write":10},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-07","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-14","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-04","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-29","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}}}},"minimax-cn-coding-plan":{"id":"minimax-cn-coding-plan","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.cn/anthropic/v1","name":"MiniMax Token Plan (minimax.cn)","doc":"https://platform.minimaxi.com/docs/token-plan/intro","models":{"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash-Preview":{"id":"MiniMax-M3.1-Flash-Preview","name":"MiniMax-M3.1-Flash-Preview","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-27","last_updated":"2026-09-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash-Preview"},"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"}}},"ambient":{"id":"ambient","env":["AMBIENT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ambient.xyz/v1","name":"Ambient","doc":"https://ambient.xyz","models":{"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131702},"cost":{"input":0.23,"output":0.96,"cache_read":0.05,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.32,"output":3.2,"cache_read":0.16,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B-A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.09,"output":0.3,"cache_read":0.09,"cache_write":0},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.32,"output":3.2,"cache_read":0.16,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":1,"cache_read":0.15,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"ambient/large":{"id":"ambient/large","name":"Ambient Large","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"cost":{"input":0.6,"output":2,"cache_read":0.15,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"ambient/small":{"id":"ambient/small","name":"Ambient Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":101072,"output":101072},"cost":{"input":0.001,"output":0.001,"cache_read":0,"cache_write":0}},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"cost":{"input":0.6,"output":2,"cache_read":0.15,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"unsloth/qwen3.8-27b-gguf":{"id":"unsloth/qwen3.8-27b-gguf","name":"qwen3.8-27b-4q","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"unsloth","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":101072,"output":101072},"cost":{"input":0.001,"output":0.001,"cache_read":0,"cache_write":0}},"zai-org/GLM-5.2-FP8":{"id":"zai-org/GLM-5.2-FP8","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"status":"deprecated","cost":{"input":1.2,"output":4.2,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"berget":{"id":"berget","env":["BERGET_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.berget.ai/v1","name":"Berget.AI","doc":"https://api.berget.ai","models":{"Qwen/Qwen3.8-27B-FP8":{"id":"Qwen/Qwen3.8-27B-FP8","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.46,"output":3.48},"canonical_model_id":"alibaba/qwen3.8-27b"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B Instruct","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["audio","image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.280375,"output":0.56075},"canonical_model_id":"google/gemma-4-31b-it"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"output":32768},"cost":{"input":3.3645,"output":16.8225,"cache_read":1.00935},"canonical_model_id":"moonshotai/kimi-k3"},"mistralai/Mistral-Small-3.2-24B-Instruct-2506":{"id":"mistralai/Mistral-Small-3.2-24B-Instruct-2506","name":"Mistral Small 3.2 24B Instruct 2506","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":8192},"cost":{"input":0.33645,"output":0.33645},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"cost":{"input":0.29,"output":0.58},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":32768},"status":"deprecated","cost":{"input":1.54,"output":4.84},"canonical_model_id":"zhipuai/glm-5.2"}}},"friendli":{"id":"friendli","env":["FRIENDLI_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://api.friendli.ai/serverless/v1","name":"Friendli","doc":"https://friendli.ai/docs/guides/serverless_endpoints/introduction","models":{"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":64000},"cost":{"input":0.5,"output":1.5,"cache_read":0.25},"canonical_model_id":"deepseek/deepseek-v3.2"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.26,"output":3.96,"cache_read":0.234},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"engy":{"id":"engy","env":["ENGY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.engy.ai/v1","name":"engy","doc":"https://engy.ai/pricing","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1001536,"input":936000,"output":65536},"cost":{"input":0.045,"output":0.32,"cache_read":0.015},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":983040,"output":65536},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"input":294912,"output":32768},"cost":{"input":0.98,"output":3.08,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1113088,"input":1047552,"output":65536},"cost":{"input":1.95,"output":9.75,"cache_read":0.195},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":983040,"output":65536},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":920576,"output":128000},"cost":{"input":0.045,"output":0.09,"cache_read":0.009},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":229376,"output":32768},"cost":{"input":0.68,"output":1.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"input":294912,"output":32768},"cost":{"input":0.135,"output":0.45,"cache_read":0.027},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":208192,"input":200000,"output":8192},"cost":{"input":0.045,"output":0.3,"cache_read":0.015},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"nvidia":{"id":"nvidia","env":["NVIDIA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://integrate.api.nvidia.com/v1","name":"Nvidia","doc":"https://docs.api.nvidia.com/nim/","models":{"black-forest-labs/flux_1-kontext-dev":{"id":"black-forest-labs/flux_1-kontext-dev","type":"image","name":"FLUX.1-Kontext-dev","description":"Open-weight FLUX.1 Kontext model for instruction-based image editing","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-29","last_updated":"2025-05-29","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.1-kontext-dev"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-1-kontext-dev"},"black-forest-labs/flux_2-klein-4b":{"id":"black-forest-labs/flux_2-klein-4b","type":"image","name":"FLUX.2 Klein 4B","description":"Fast 4B open-weight image generation and editing model","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-06","release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.2-klein-4b"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-2-klein-4b"},"black-forest-labs/flux_1-schnell":{"id":"black-forest-labs/flux_1-schnell","type":"image","name":"FLUX.1-schnell","description":"Fast few-step FLUX.1 text-to-image model under Apache 2.0","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-07","release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.1-schnell"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-1-schnell"},"black-forest-labs/flux.1-dev":{"id":"black-forest-labs/flux.1-dev","type":"image","name":"FLUX.1-dev","description":"Open-weight FLUX.1 text-to-image model for high-quality guidance-distilled generation","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-08","release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.1-dev"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-1-dev"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-3-12b-it"},"google/diffusiongemma-26b-a4b-it":{"id":"google/diffusiongemma-26b-a4b-it","name":"DiffusionGemma 26B A4B IT","description":"Gemma 4 based discrete text diffusion model for fast chat, reasoning, and tool use","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":250000,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/diffusiongemma-26b-a4b-it"},"google/google-paligemma":{"id":"google/google-paligemma","name":"paligemma","description":"Vision-language model for answering questions about images and text","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-05-14","last_updated":"2024-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma-4-31B-IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"openai/whisper-large-v3":{"id":"openai/whisper-large-v3","type":"transcription","name":"Whisper Large v3","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-09","release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"openai/whisper-large-v3"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-20b"},"deepseek-ai/deepseek-v4.1-flash":{"id":"deepseek-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"stepfun-ai/step-5-preview":{"id":"stepfun-ai/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":1048576},"cost":{"input":0,"output":0},"canonical_model_id":"stepfun/step-5-preview"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"meta/llama-3.2-11b-vision-instruct":{"id":"meta/llama-3.2-11b-vision-instruct","name":"Llama 3.2 11b Vision Instruct","description":"Open multimodal Llama model for image understanding, captioning, and visual QA","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"meta/llama-guard-4-12b":{"id":"meta/llama-guard-4-12b","name":"Llama Guard 4 12B","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-guard-4-12b"},"meta/llama-3.2-90b-vision-instruct":{"id":"meta/llama-3.2-90b-vision-instruct","name":"Llama-3.2-90B-Vision-Instruct","description":"Large open multimodal Llama model for image reasoning, document QA, and captioning","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.2-90b-vision-instruct"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"meta/muse-glimmer-30b"},"mistralai/magistral-small-2506":{"id":"mistralai/magistral-small-2506","name":"Magistral Small 2506","description":"Open Mistral reasoning model for transparent step-by-step problem solving","family":"magistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/magistral-small-2506"},"mistralai/mistral-medium-3-instruct":{"id":"mistralai/mistral-medium-3-instruct","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/mistral-7b-instruct-v0.3":{"id":"mistralai/mistral-7b-instruct-v0.3","name":"Mistral-7B-Instruct-v0.3","description":"Open 7B Mistral instruction model with function calling","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-22","last_updated":"2024-05-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-7b-instruct-v0.3"},"qwen/qwen-image":{"id":"qwen/qwen-image","type":"image","name":"Qwen Image","description":"Text-to-image foundation model with multilingual text rendering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0}},"qwen/qwen-image-edit":{"id":"qwen/qwen-image-edit","type":"image","name":"Qwen Image Edit","description":"Image editing model with multilingual text editing and subject consistency","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-08-19","last_updated":"2025-08-19","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0}},"minimaxai/minimax-m3.1-flash":{"id":"minimaxai/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"nvidia/riva-translate-4b-instruct-v1.1":{"id":"nvidia/riva-translate-4b-instruct-v1.1","name":"riva-translate-4b-instruct-v1_1","description":"Translation model for multilingual conversion, localization, and cross-language workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-12","last_updated":"2025-12-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/usdcode":{"id":"nvidia/usdcode","name":"usdcode","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-01","last_updated":"2026-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/sparsedrive":{"id":"nvidia/sparsedrive","name":"sparsedrive","description":"End-to-end autonomous driving stack for perception, prediction, and planning","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-03-18","last_updated":"2025-07-20","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"nvidia/nemotron-3.5-lightning-30b-a3b":{"id":"nvidia/nemotron-3.5-lightning-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/llama-nemotron-embed-vl-1b-v2":{"id":"nvidia/llama-nemotron-embed-vl-1b-v2","type":"embedding","name":"llama-nemotron-embed-vl-1b-v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"nemotron","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-nemotron-embed-vl-1b-v2"},"nvidia/nemotron-voicechat":{"id":"nvidia/nemotron-voicechat","name":"nemotron-voicechat","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-voicechat"},"nvidia/llama-3.1-nemotron-70b-instruct":{"id":"nvidia/llama-3.1-nemotron-70b-instruct","name":"Llama 3.1 Nemotron 70B Instruct","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-3.1-nemotron-70b-instruct"},"nvidia/rerank-qa-mistral-4b":{"id":"nvidia/rerank-qa-mistral-4b","type":"reranking","name":"rerank-qa-mistral-4b","description":"Reranking model for improving retrieval quality in search and recommendation systems","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-17","last_updated":"2025-01-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"api":"https://ai.api.nvidia.com/v1/retrieval/nvidia/reranking"},"cost":{"input":0,"output":0}},"nvidia/nemotron-3.5-super-vl-120b-a12b":{"id":"nvidia/nemotron-3.5-super-vl-120b-a12b","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/llama-nemotron-rerank-vl-1b-v2":{"id":"nvidia/llama-nemotron-rerank-vl-1b-v2","type":"reranking","name":"llama-nemotron-rerank-vl-1b-v2","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"nemotron","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10240,"output":4096},"provider":{"api":"https://ai.api.nvidia.com/v1/retrieval/nvidia/llama-nemotron-rerank-vl-1b-v2/reranking"},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-nemotron-rerank-vl-1b-v2"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]},{"type":"budget_tokens","min":-1,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/active-speaker-detection":{"id":"nvidia/active-speaker-detection","name":"Active Speaker Detection","description":"Detects and tracks which person is speaking across video frames","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/cosmos-reason2-8b":{"id":"nvidia/cosmos-reason2-8b","name":"Cosmos Reason2 8B","description":"Vision language model for physical-world understanding with structured reasoning on video and images","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning":{"id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning","name":"Nemotron 3 Nano Omni","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":-1,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/studiovoice":{"id":"nvidia/studiovoice","name":"studiovoice","description":"Enhances noisy or reverberant speech recordings to studio-quality speech","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-03","last_updated":"2025-06-13","modalities":{"input":["audio"],"output":["audio"]},"open_weights":true,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0}},"nvidia/cosmos-transfer2_5-2b":{"id":"nvidia/cosmos-transfer2_5-2b","type":"video","name":"cosmos-transfer2.5-2b","description":"Generates physics-aware video world states from text and spatial control videos","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/bevformer":{"id":"nvidia/bevformer","name":"bevformer","description":"Multi-frame bird's-eye-view 3D perception transformer for autonomous driving","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-03-18","last_updated":"2025-07-20","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"nvidia/streampetr":{"id":"nvidia/streampetr","name":"streampetr","description":"Temporal 3D object detection model for autonomous driving","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"nvidia/magpie-tts-zeroshot":{"id":"nvidia/magpie-tts-zeroshot","type":"speech","name":"magpie-tts-zeroshot","description":"Expressive text-to-speech that clones a voice from a short audio sample","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-22","last_updated":"2025-06-12","modalities":{"input":["text","audio"],"output":["audio"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/synthetic-video-detector":{"id":"nvidia/synthetic-video-detector","name":"synthetic-video-detector","description":"Detects AI-generated (synthetic) video","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/llama-3.1-nemotron-safety-guard-8b-v3":{"id":"nvidia/llama-3.1-nemotron-safety-guard-8b-v3","name":"llama-3.1-nemotron-safety-guard-8b-v3","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-3.1-nemotron-safety-guard-8b-v3"},"nvidia/usdvalidate":{"id":"nvidia/usdvalidate","name":"usdvalidate","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-07-24","last_updated":"2025-01-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":4096},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/llama-3.1-nemotron-ultra-253b-v1":{"id":"nvidia/llama-3.1-nemotron-ultra-253b-v1","name":"Llama 3.1 Nemotron Ultra 253B","description":"Flagship Nemotron model for high-throughput reasoning and complex agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-04-07","last_updated":"2025-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-3.1-nemotron-ultra-253b"}}},"opencode-go":{"id":"opencode-go","env":["OPENCODE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://opencode.ai/inference/go/openai/v1","name":"OpenCode Go","doc":"https://opencode.ai/v2/docs/console/go/","models":{"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for coding and agentic workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro (New)","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"cat-jam-free":{"id":"cat-jam-free","name":"Cat Jam Free","description":"A text-only reasoning model for coding and agentic workflows, with tool calling and structured JSON outputs.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"longcat-2.0":{"id":"longcat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.0"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo V2.5","description":"MiMo omni model for text, image, video, audio, and agents","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"longcat-2.5-preview-free":{"id":"longcat-2.5-preview-free","name":"LongCat 2.5 Preview Free","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meituan/longcat-2.5-preview"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter multimodal flagship for coding, professional work, and long-horizon agentic workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax-m3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"space-bunny":{"id":"space-bunny","name":"Space Bunny","description":"Anonymous preview reasoning model for coding, agentic tasks, tool use, and multimodal input","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":524288,"output":524288},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.03,"cache_write":0}},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"step-5-preview-free":{"id":"step-5-preview-free","name":"Step 5 Preview Free","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"stepfun/step-5-preview"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"minimax-m3.1-flash-free":{"id":"minimax-m3.1-flash-free","name":"MiniMax M3.1 Flash Free","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen3.7-plus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.4,"output":1.6,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"beam-free":{"id":"beam-free","name":"Beam Free","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"muse-spark-1.2-contributor":{"id":"muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"},"gemini-4-argon-contributor":{"id":"gemini-4-argon-contributor","name":"Gemini 4 Argon Contributor","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/go/google/v1beta"},"cost":{"input":4,"output":20,"cache_read":0.2,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"}}},"anyapi":{"id":"anyapi","env":["ANYAPI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.anyapi.ai/v1","name":"AnyAPI","doc":"https://docs.anyapi.ai","models":{"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"canonical_model_id":"xai/grok-4.3"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-r1":{"id":"deepseek/deepseek-r1","name":"DeepSeek Reasoner","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-reasoner"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"canonical_model_id":"cohere/command-r-plus-08-2024"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3-pro-preview":{"id":"google/gemini-3-pro-preview","name":"Gemini 3 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3-pro-preview"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-2.5-flash-lite"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.1"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o4-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o3"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1"},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"canonical_model_id":"perplexity/sonar-pro"},"perplexity/sonar-reasoning-pro":{"id":"perplexity/sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"canonical_model_id":"mistral/mistral-large-2512"},"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","canonical_model_id":"mistral/devstral-2512"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"canonical_model_id":"anthropic/claude-sonnet-4-6"}}},"modelscope":{"id":"modelscope","env":["MODELSCOPE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api-inference.modelscope.cn/v1","name":"ModelScope","doc":"https://modelscope.cn/docs/model-service/API-Inference/intro","models":{"ZhipuAI/GLM-4.6":{"id":"ZhipuAI/GLM-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.6"},"ZhipuAI/GLM-4.5":{"id":"ZhipuAI/GLM-4.5","name":"GLM-4.5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.5"},"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3-235B-A22B-Thinking-2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3-30B-A3B-Thinking-2507":{"id":"Qwen/Qwen3-30B-A3B-Thinking-2507","name":"Qwen3 30B A3B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"}}},"merge-gateway":{"id":"merge-gateway","env":["MERGE_GATEWAY_API_KEY"],"npm":"merge-gateway-ai-sdk-provider","api":"https://api-gateway.merge.dev/v1/ai-sdk","name":"Merge Gateway","doc":"https://docs.merge.dev/merge-gateway","models":{"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":2,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash-0731-fast":{"id":"deepseek/deepseek-v4-flash-0731-fast","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.1125},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.035,"output":0.07,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash-0423":{"id":"deepseek/deepseek-v4-flash-0423","name":"DeepSeek V4 Flash 0423","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.139,"output":0.278},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-v3":{"id":"deepseek/deepseek-v3","name":"DeepSeek V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":81920},"cost":{"input":0.27,"output":1.1},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek/deepseek-v4-pro-0423":{"id":"deepseek/deepseek-v4-pro-0423","name":"DeepSeek V4 Pro 0423","description":"DeepSeek V4 Pro initial snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.65,"output":3.3},"canonical_model_id":"deepseek/deepseek-v4-pro-0423"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 (0528)","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":32768},"cost":{"input":1.35,"output":5.4,"cache_read":0.945},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.56,"output":1.68,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek/deepseek-r1":{"id":"deepseek/deepseek-r1","name":"DeepSeek R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":40960},"cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.28,"output":0.4,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/sakana-namazu":{"id":"sakana/sakana-namazu","name":"Sakana Namazu","description":"Japanese-specialized reasoning model based on Kimi K2.6 and tuned for Japanese language, culture, and business workflows","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.7,"output":2.2,"cache_read":0.13},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"Glm 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.7-flash":{"id":"zai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM-4.7 FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":50000}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.05,"output":3.3,"cache_read":0.195},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"output":32000},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax M2.5 Highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":128000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 Highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+ 08-2024","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B 12-2024","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R 08-2024","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"cohere/command-a-03-2025":{"id":"cohere/command-a-03-2025","name":"Command A 03-2025","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-03-2025"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.08,"output":0.45,"cache_read":0.04},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-flash-latest"},"google/gemini-embedding-001":{"id":"google/gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":4096},"cost":{"input":0.15,"output":0},"canonical_model_id":"google/gemini-embedding-001"},"google/gemini-2.5-computer-use-preview-10-2025":{"id":"google/gemini-2.5-computer-use-preview-10-2025","name":"Gemini 2.5 Computer Use Preview (10-2025)","description":"Specialized Gemini 2.5 model for browser-control agents that automate UI tasks","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-10-07","last_updated":"2025-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":1.25,"output":10},"canonical_model_id":"google/gemini-2.5-computer-use-preview-10-2025"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash-Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Gemini 3 Pro Image","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Gemini 3.1 Flash Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B It","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-flash-lite-latest"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Gemini 2.5 Flash Image","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-embedding-2":{"id":"google/gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.2,"output":0},"canonical_model_id":"google/gemini-embedding-2"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"mistral/mistral-medium-latest":{"id":"mistral/mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-latest"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/devstral-2512":{"id":"mistral/devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistral/codestral-latest":{"id":"mistral/codestral-latest","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-latest"},"mistral/mistral-small-latest":{"id":"mistral/mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-latest"},"mistral/magistral-medium-latest":{"id":"mistral/magistral-medium-latest","name":"Magistral Medium (latest)","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral-medium","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-03-17","last_updated":"2025-03-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2,"output":5},"canonical_model_id":"mistral/magistral-medium-latest"},"mistral/mistral-medium-2505":{"id":"mistral/mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-2505"},"mistral/pixtral-large-latest":{"id":"mistral/pixtral-large-latest","name":"Pixtral Large (latest)","description":"Mistral's larger vision model for document-heavy image understanding and chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2024-11-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-latest"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/devstral-medium-latest":{"id":"mistral/devstral-medium-latest","name":"Devstral 2 (latest)","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-medium-latest"},"openai/gpt-5-search-api":{"id":"openai/gpt-5-search-api","type":"chat","name":"GPT-5 Search API","description":"GPT-5 model trained for web search in the Chat Completions API","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5-search-api"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-audio-mini":{"id":"openai/gpt-audio-mini","type":"chat","name":"GPT Audio Mini","description":"Cost-efficient audio model for speech input and output with the Chat Completions API","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":2.4},"canonical_model_id":"openai/gpt-audio-mini"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-oss-safeguard-120b":{"id":"openai/gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.2-chat-latest":{"id":"openai/gpt-5.2-chat-latest","name":"GPT-5.2 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3 Mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT-OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.04,"output":0.15,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.07,"output":0.2,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.3-chat-latest":{"id":"openai/gpt-5.3-chat-latest","name":"GPT-5.3 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o Mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.1-chat-latest":{"id":"openai/gpt-5.1-chat-latest","name":"GPT-5.1 Chat Latest","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5-chat-latest":{"id":"openai/gpt-5-chat-latest","name":"GPT-5 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-chat-latest"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4 Mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-audio":{"id":"openai/gpt-audio","type":"chat","name":"GPT Audio","description":"Audio model for speech input and output with the Chat Completions API","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-audio"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT-OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.05,"output":0.25},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/whisper-1":{"id":"openai/whisper-1","type":"transcription","name":"Whisper","description":"General-purpose speech recognition model for transcription and translation","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2023-03-01","last_updated":"2023-03-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":448000,"output":4096},"provider":{"npm":"@ai-sdk/openai","api":"https://api-gateway.merge.dev/v1"},"canonical_model_id":"openai/whisper-1"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"bytedance/dola-seed-2.0-mini":{"id":"bytedance/dola-seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance/dola-seed-2.0-pro":{"id":"bytedance/dola-seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"bytedance/dola-seed-2.0-code":{"id":"bytedance/dola-seed-2.0-code","name":"Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4,"output":2.4},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"bytedance/dola-seed-2.0-lite":{"id":"bytedance/dola-seed-2.0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.25,"output":2},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance/dola-seed-2.0-code-preview":{"id":"bytedance/dola-seed-2.0-code-preview","name":"Dola Seed 2.0 Code (preview)","description":"Preview coding model for repository understanding, refactors, and engineering tasks","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3}},"meta/llama-3.3-70b-instruct":{"id":"meta/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.22,"output":0.5,"cache_read":0.11},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta/llama-3.1-8b-instruct":{"id":"meta/llama-3.1-8b-instruct","name":"Llama 3.1 8B","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"meta/llama-3.1-70b-instruct":{"id":"meta/llama-3.1-70b-instruct","name":"Llama 3.1 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"cost":{"input":0.99,"output":0.99},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.5":{"id":"moonshot/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3.1-pro":{"id":"moonshot/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"qwen/qwen3-vl-235b-a22b-thinking":{"id":"qwen/qwen3-vl-235b-a22b-thinking","name":"Qwen3-VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.287,"output":2.867,"cache_read":0.0574},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3-VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.21,"output":1.9,"cache_read":0.105},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.276,"output":1.651,"cache_read":0.0276,"cache_write":0.345},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-235b-a22b":{"id":"qwen/qwen3-235b-a22b","name":"Qwen3 235B A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.287,"output":1.147,"cache_read":0.0574},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"qwen/qwen3.5-4b":{"id":"qwen/qwen3.5-4b","name":"Qwen3.5 4B","description":"Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.04,"output":0.07},"canonical_model_id":"alibaba/qwen3.5-4b"},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.144,"output":0.574,"cache_read":0.0288,"cache_write":0.18},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.825,"output":2.4755,"cache_read":0.165,"cache_write":1.03125},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3-235b-a22b-instruct-2507":{"id":"qwen/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.0574},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.057,"output":0.459,"cache_read":0.020357},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-vl-plus":{"id":"qwen/qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.143,"output":1.434,"cache_read":0.0286,"cache_write":0.17875},"canonical_model_id":"alibaba/qwen3-vl-plus"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.574,"output":2.294,"cache_read":0.1148,"cache_write":0.7175},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen3-vl-flash":{"id":"qwen/qwen3-vl-flash","name":"Qwen3-VL Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.022,"output":0.215,"cache_read":0.0044,"cache_write":0.0275},"canonical_model_id":"alibaba/qwen3-vl-flash"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.359,"output":1.434,"cache_read":0.0718,"cache_write":0.44875},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen-flash":{"id":"qwen/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.022,"output":0.216,"cache_read":0.0044,"cache_write":0.0275},"canonical_model_id":"alibaba/qwen-flash"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":0.13},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.115,"output":0.917,"cache_read":0.023},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.6-max-preview":{"id":"qwen/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1.31,"output":7.88},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.172,"output":1.032,"cache_read":0.0344},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":1010000},"cost":{"input":2.5,"output":6.25,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.115,"output":0.688,"cache_read":0.0115,"cache_write":0.14375},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3-max-preview":{"id":"qwen/qwen3-max-preview","name":"Qwen3 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-05","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1.08,"output":4.8},"canonical_model_id":"alibaba/qwen3-max-preview"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.4126,"output":2.4754},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":977000,"output":128000},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.115,"output":0.287,"cache_read":0.023,"cache_write":0.14375},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.108,"output":1.076,"cache_read":0.0216},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3-VL 30B-A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.0216},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.029,"output":0.287,"cache_read":0.0029,"cache_write":0.03625},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3-coder-480b-a35b-instruct":{"id":"qwen/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.22,"output":1.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.086,"output":0.688,"cache_read":0.0172},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.15,"output":0.8,"cache_read":0.075},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":1.1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485,"cache_read":0.0496},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.165,"output":0.99,"cache_read":0.0165,"cache_write":0.20625},"canonical_model_id":"alibaba/qwen3.6-flash"},"writer/palmyra-x4":{"id":"writer/palmyra-x4","name":"Palmyra X4","description":"Enterprise language model for workflow automation, coding, data analysis, and tool use","family":"palmyra","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-10-09","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":2.5,"output":10},"canonical_model_id":"writer/palmyra-x4"},"writer/palmyra-x5":{"id":"writer/palmyra-x5","name":"Palmyra X5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"palmyra","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"anthropic/claude-haiku-4-5-20251001":{"id":"anthropic/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5 (20251001)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (20251101)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-sonnet-4-5-20250929":{"id":"anthropic/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5 (20250929)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4-1-20250805":{"id":"anthropic/claude-opus-4-1-20250805","name":"Claude Opus 4.1 (20250805)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":128000}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-3.5-lightning-30b-a3b":{"id":"nvidia/nemotron-3.5-lightning-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-nano-9b-v2":{"id":"nvidia/nemotron-nano-9b-v2","name":"Nemotron Nano 9B","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-09","release_date":"2025-08-18","last_updated":"2025-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.06,"output":0.23},"canonical_model_id":"nvidia/nemotron-nano-9b-v2"}}},"alibaba-token-plan":{"id":"alibaba-token-plan","env":["ALIBABA_TOKEN_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1","name":"Alibaba Token Plan","doc":"https://www.alibabacloud.com/help/en/model-studio/token-plan-overview","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"qwen-image-2.0-pro":{"id":"qwen-image-2.0-pro","type":"image","name":"Qwen Image 2.0 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-plus"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"happyhorse-1.1-t2v":{"id":"happyhorse-1.1-t2v","type":"video","name":"HappyHorse 1.1 Text-to-Video","description":"Video model for prompt-driven text-to-video generation","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"qwen3.8-max-preview":{"id":"qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"wan2.7-image-pro":{"id":"wan2.7-image-pro","type":"image","name":"Wan2.7 Image Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-max"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"wan2.7-image":{"id":"wan2.7-image","type":"image","name":"Wan2.7 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"input":196601,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M2.5"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"happyhorse-1.1-r2v":{"id":"happyhorse-1.1-r2v","type":"video","name":"HappyHorse 1.1 Reference-to-Video","description":"Video model for reference-guided video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"qwen-image-2.0":{"id":"qwen-image-2.0","type":"image","name":"Qwen Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"happyhorse-1.1-i2v":{"id":"happyhorse-1.1-i2v","type":"video","name":"HappyHorse 1.1 Image-to-Video","description":"Video model for image-to-video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-03","last_updated":"2025-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0,"output":0}},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-flash"}}},"xiaomi-token-plan-ams":{"id":"xiaomi-token-plan-ams","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan-ams.xiaomimimo.com/v1","name":"Xiaomi Token Plan (Europe)","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.5-tts-voicedesign":{"id":"mimo-v2.5-tts-voicedesign","type":"speech","name":"MiMo-V2.5-TTS-VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voicedesign"},"mimo-v2.5-tts-voiceclone":{"id":"mimo-v2.5-tts-voiceclone","type":"speech","name":"MiMo-V2.5-TTS-VoiceClone","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voiceclone"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo-v2.5-tts":{"id":"mimo-v2.5-tts","type":"speech","name":"MiMo-V2.5-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts"}}},"echo":{"id":"echo","env":["ECHO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://echo.tracerml.ai/v1","name":"Echo","doc":"https://echo.tracerml.ai/docs/api","models":{"echo":{"id":"echo","name":"Echo","description":"Adaptive model for coding, reasoning, and tool-driven agent workflows through one OpenAI-compatible endpoint","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"beta","cost":{"input":10,"output":50}}}},"coralbricks":{"id":"coralbricks","env":["CORAL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.coralbricks.ai/v1","name":"CoralBricks","doc":"https://www.coralbricks.ai/docs","models":{"deepseek-v4.1-flash-fast":{"id":"deepseek-v4.1-flash-fast","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.25,"output":1.2,"cache_read":0.005,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"}}},"jiekou":{"id":"jiekou","env":["JIEKOU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.jiekou.ai/openai","name":"Jiekou.AI","doc":"https://docs.jiekou.ai/docs/support/quickstart?utm_source=github_models.dev","models":{"gemini-2.5-flash-lite-preview-06-17":{"id":"gemini-2.5-flash-lite-preview-06-17","name":"gemini-2.5-flash-lite-preview-06-17","description":"First preview of Gemini 2.5 Flash-Lite, the low-cost, low-latency 2.5 model","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.09,"output":0.36},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-06-17"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"claude-haiku-4-5-20251001","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"gpt-5.1-codex-mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2375,"output":1.9,"cache_read":0.0237},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"grok-4-0709":{"id":"grok-4-0709","name":"grok-4-0709","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-01","last_updated":"2026-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":2.85,"output":14.25}},"gpt-5.2-pro":{"id":"gpt-5.2-pro","name":"gpt-5.2-pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":19.95,"output":159.6},"canonical_model_id":"openai/gpt-5.2-pro"},"gemini-2.5-flash-lite-preview-09-2025":{"id":"gemini-2.5-flash-lite-preview-09-2025","name":"gemini-2.5-flash-lite-preview-09-2025","description":"September 2025 preview of Gemini 2.5 Flash-Lite with better instruction following and multimodal quality","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.09,"output":0.36},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-09-2025"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"gpt-5.2-codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"claude-opus-4-5-20251101","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":65536},"cost":{"input":4.75,"output":23.75,"cache_read":0.475,"cache_write":5.9375},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"gemini-2.5-pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"google/gemini-2.5-pro"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"grok-code-fast-1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":10000},"cost":{"input":0.19,"output":1.425,"cache_read":0.019},"canonical_model_id":"xai/grok-code-fast-1"},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.045,"output":4.18,"cache_read":0.5225},"canonical_model_id":"openai/o3-mini"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"claude-sonnet-4-5-20250929","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-sonnet-4-20250514":{"id":"claude-sonnet-4-20250514","name":"claude-sonnet-4-20250514","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":2.7,"output":13.5},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"gpt-5.1-codex-max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5.1-codex-max"},"claude-opus-4-20250514":{"id":"claude-opus-4-20250514","name":"claude-opus-4-20250514","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":13.5,"output":67.5},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"gpt-5.2":{"id":"gpt-5.2","name":"gpt-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.6625,"output":13.3,"cache_read":0.1662},"canonical_model_id":"openai/gpt-5.2"},"gpt-5-mini":{"id":"gpt-5-mini","name":"gpt-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2375,"output":1.9,"cache_read":0.0237},"canonical_model_id":"openai/gpt-5-mini"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"claude-opus-4-1-20250805","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":13.5,"output":67.5},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gpt-5-pro":{"id":"gpt-5-pro","name":"gpt-5-pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":14.25,"output":114},"canonical_model_id":"openai/gpt-5-pro"},"grok-4-fast-reasoning":{"id":"grok-4-fast-reasoning","name":"grok-4-fast-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","cost":{"input":0.18,"output":0.45},"canonical_model_id":"xai/grok-4-fast-reasoning"},"gpt-5.1":{"id":"gpt-5.1","name":"gpt-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5.1"},"gpt-5-chat-latest":{"id":"gpt-5-chat-latest","name":"gpt-5-chat-latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.125,"output":9},"canonical_model_id":"openai/gpt-5-chat-latest"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"claude-opus-4-6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gemini-2.5-pro-preview-06-05":{"id":"gemini-2.5-pro-preview-06-05","name":"gemini-2.5-pro-preview-06-05","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["image","text","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"google/gemini-2.5-pro-preview"},"gemini-2.5-flash-preview-05-20":{"id":"gemini-2.5-flash-preview-05-20","name":"gemini-2.5-flash-preview-05-20","description":"Preview snapshot of Gemini 2.5 Flash with controllable thinking budgets","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1425,"output":3.325,"cache_read":0.0285},"canonical_model_id":"google/gemini-2.5-flash-preview-05-20"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"gemini-3-flash-preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.475,"output":2.85,"cache_read":0.0475},"canonical_model_id":"google/gemini-3-flash-preview"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"grok-4-1-fast-non-reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.19,"output":0.475},"canonical_model_id":"xai/grok-4.1-fast"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4},"canonical_model_id":"openai/o4-mini"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"gemini-2.5-flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.285,"output":2.375,"cache_read":0.0285},"canonical_model_id":"google/gemini-2.5-flash"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"grok-4-1-fast-reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.19,"output":0.475},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"gpt-5-nano":{"id":"gpt-5-nano","name":"gpt-5-nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.0475,"output":0.38,"cache_read":0.0047},"canonical_model_id":"openai/gpt-5-nano"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"gpt-5.1-codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5.1-codex"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":9.5,"output":38,"cache_read":0.475},"canonical_model_id":"openai/o3"},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"grok-4-fast-non-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","cost":{"input":0.18,"output":0.45},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"gpt-5-codex":{"id":"gpt-5-codex","name":"gpt-5-codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5-codex"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"gemini-3-pro-preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.8,"output":10.8},"canonical_model_id":"google/gemini-3-pro-preview"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"gemini-2.5-flash-lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.095,"output":0.38,"cache_read":0.0095},"canonical_model_id":"google/gemini-2.5-flash-lite"},"deepseek/deepseek-v3-0324":{"id":"deepseek/deepseek-v3-0324","name":"DeepSeek V3 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.28,"output":1.14,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.7,"output":2.5,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.27,"output":1},"canonical_model_id":"deepseek/deepseek-v3.1"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"Minimax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"baidu/ernie-4.5-vl-424b-a47b":{"id":"baidu/ernie-4.5-vl-424b-a47b","name":"ERNIE 4.5 VL 424B A47B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":123000,"output":16000},"cost":{"input":0.42,"output":1.25},"canonical_model_id":"baidu/ernie-4.5-vl-424b-a47b"},"baidu/ernie-4.5-300b-a47b-paddle":{"id":"baidu/ernie-4.5-300b-a47b-paddle","name":"ERNIE 4.5 300B A47B","description":"Open ERNIE text MoE with 300B total and 47B active parameters","family":"ernie","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":123000,"output":12000},"cost":{"input":0.28,"output":1.1},"canonical_model_id":"baidu/ernie-4.5-300b-a47b"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2-instruct":{"id":"moonshotai/kimi-k2-instruct","name":"Kimi K2 Instruct","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":100352},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":100352},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B Thinking","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3-235b-a22b-instruct-2507":{"id":"qwen/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.15,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22b Thinking 2507","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":3},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen/qwen3-30b-a3b-fp8":{"id":"qwen/qwen3-30b-a3b-fp8","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.09,"output":0.45},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen3-235b-a22b-fp8":{"id":"qwen/qwen3-235b-a22b-fp8","name":"Qwen3 235B A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"qwen/qwen3-32b-fp8":{"id":"qwen/qwen3-32b-fp8","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.1,"output":0.45},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3-coder-480b-a35b-instruct":{"id":"qwen/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"qwen/qwen3-coder-next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"xiaomimimo/mimo-v2-flash":{"id":"xiaomimimo/mimo-v2-flash","name":"XiaomiMiMo/MiMo-V2-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"xiaomi/mimo-v2-flash"},"minimaxai/minimax-m1-80k":{"id":"minimaxai/minimax-m1-80k","name":"MiniMax M1","description":"Open hybrid-attention MoE reasoning model with a million-token context and an 80K thinking budget","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":40000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"minimax/minimax-m1-80k"},"zai-org/glm-4.5v":{"id":"zai-org/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8},"canonical_model_id":"zhipuai/glm-4.5v"},"zai-org/glm-4.7-flash":{"id":"zai-org/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/glm-4.7":{"id":"zai-org/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/glm-4.5":{"id":"zai-org/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.5"}}},"google-vertex":{"id":"google-vertex","env":["GOOGLE_VERTEX_PROJECT","GOOGLE_VERTEX_LOCATION","GOOGLE_APPLICATION_CREDENTIALS"],"npm":"@ai-sdk/google-vertex","name":"Vertex","doc":"https://cloud.google.com/vertex-ai/generative-ai/docs/models","models":{"claude-sonnet-4@20250514":{"id":"claude-sonnet-4@20250514","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"claude-fable-5-1@default":{"id":"claude-fable-5-1@default","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gemini-embedding-001":{"id":"gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"cost":{"input":0.15,"output":0},"canonical_model_id":"google/gemini-embedding-001"},"claude-opus-4-7@default":{"id":"claude-opus-4-7@default","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"claude-sonnet-5@default":{"id":"claude-sonnet-5@default","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"gemini-3.1-pro-preview-customtools":{"id":"gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"beta","cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"claude-opus-4-5@20251101":{"id":"claude-opus-4-5@20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"claude-sonnet-4-6@default":{"id":"claude-sonnet-4-6@default","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"mistral-medium-3":{"id":"mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/mistral","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/publishers/mistralai/models"},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-2505"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"claude-opus-4@20250514":{"id":"claude-opus-4@20250514","name":"Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"claude-sonnet-4-5@20250929":{"id":"claude-sonnet-4-5@20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-opus-4-8@default":{"id":"claude-opus-4-8@default","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-opus-4-6@default":{"id":"claude-opus-4-6@default","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"claude-opus-5-5@default":{"id":"claude-opus-5-5@default","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gemini-nano-banana-2.1":{"id":"gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Image model for prompt-driven generation, multi-turn editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":1.5,"output":30,"cache_read":0.15},"canonical_model_id":"google/gemini-nano-banana-2.1"},"claude-opus-4-1@20250805":{"id":"claude-opus-4-1@20250805","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"beta","cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"beta","cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"mistral-small-2503":{"id":"mistral-small-2503","name":"Mistral Small 3.1 24B","description":"Efficient multimodal model for instruction following, coding, reasoning, and function calling","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/mistral","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/publishers/mistralai/models"},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"gemini-2.5-flash-tts":{"id":"gemini-2.5-flash-tts","type":"speech","name":"Gemini 2.5 Flash TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-09-30","last_updated":"2025-12-10","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":32768,"input":8192,"output":16384},"cost":{"input":0.5,"output":10},"canonical_model_id":"google/gemini-2.5-flash-tts"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":120,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1}},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":60,"cache_read":0.05},"canonical_model_id":"google/gemini-3.1-flash-image"},"claude-sonnet-5-5@default":{"id":"claude-sonnet-5-5@default","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"claude-haiku-4-5@20251001":{"id":"claude-haiku-4-5@20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gemini-3.1-flash-lite-image":{"id":"gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":30,"cache_read":0.025},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"claude-haiku-5-5@default":{"id":"claude-haiku-5-5@default","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"claude-fable-5@default":{"id":"claude-fable-5@default","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"codestral-2":{"id":"codestral-2","name":"Codestral 2","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"provider":{"npm":"@ai-sdk/mistral","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/publishers/mistralai/models"},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-2508"},"gemini-2.5-pro-tts":{"id":"gemini-2.5-pro-tts","type":"speech","name":"Gemini 2.5 Pro TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-pro","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-09-30","last_updated":"2025-12-10","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":32768,"input":8192,"output":16384},"cost":{"input":1,"output":20},"canonical_model_id":"google/gemini-2.5-pro-tts"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30},"canonical_model_id":"google/gemini-2.5-flash-image"},"claude-opus-5@default":{"id":"claude-opus-5@default","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"xai/grok-4.1-fast-non-reasoning":{"id":"xai/grok-4.1-fast-non-reasoning","name":"Grok 4.1 Fast","description":"Fast Grok model for responsive chat, tool-assisted work, and low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":30000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":30000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":500000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.20-reasoning":{"id":"xai/grok-4.20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":500000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.1-fast-reasoning":{"id":"xai/grok-4.1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":30000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"xai/grok-4.20-non-reasoning":{"id":"xai/grok-4.20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"google/gemma-4-26b-a4b-it-maas":{"id":"google/gemma-4-26b-a4b-it-maas","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"status":"alpha","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"openai/gpt-oss-120b-maas":{"id":"openai/gpt-oss-120b-maas","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.09,"output":0.36},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b-maas":{"id":"openai/gpt-oss-20b-maas","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.07,"output":0.25,"cache_read":0.007}},"deepseek-ai/deepseek-v3.1-maas":{"id":"deepseek-ai/deepseek-v3.1-maas","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.6,"output":1.7,"cache_read":0.06}},"deepseek-ai/deepseek-v3.2-maas":{"id":"deepseek-ai/deepseek-v3.2-maas","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-12-17","last_updated":"2026-04-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.56,"output":1.68,"cache_read":0.056}},"moonshotai/kimi-k2-thinking-maas":{"id":"moonshotai/kimi-k2-thinking-maas","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.6,"output":2.5,"cache_read":0.06}},"meta/llama-3.3-70b-instruct-maas":{"id":"meta/llama-3.3-70b-instruct-maas","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.72,"output":0.72}},"meta/llama-4-maverick-17b-128e-instruct-maas":{"id":"meta/llama-4-maverick-17b-128e-instruct-maas","name":"Llama 4 Maverick 17B 128E Instruct","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":8192},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.35,"output":1.15}},"meta/llama-4-scout-17b-16e-instruct-maas":{"id":"meta/llama-4-scout-17b-16e-instruct-maas","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1310720,"output":8192},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.25,"output":0.7},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"qwen/qwen3-235b-a22b-instruct-2507-maas":{"id":"qwen/qwen3-235b-a22b-instruct-2507-maas","name":"Qwen3 235B A22B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-13","last_updated":"2025-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.22,"output":0.88}},"zai-org/glm-5-maas":{"id":"zai-org/glm-5-maas","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1,"output":3.2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"zai-org/glm-5.2-maas":{"id":"zai-org/glm-5.2-maas","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-4.7-maas":{"id":"zai-org/glm-4.7-maas","name":"GLM-4.7","description":"GLM model for coding, tool use, and complex reasoning","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-06","last_updated":"2026-01-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.6,"output":2.2,"cache_read":0.06}}}},"salad-cloud":{"id":"salad-cloud","env":["SALAD_CLOUD_API_KEY"],"npm":"@saladtechnologies-oss/ai-sdk-provider","name":"SaladCloud AI Gateway","doc":"https://docs.salad.com/ai-gateway/explanation/overview","models":{"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Qwen MoE for agentic tasks, complex reasoning, code generation, and instruction following","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.09,"output":0.6},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"databricks":{"id":"databricks","env":["DATABRICKS_HOST","DATABRICKS_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://${DATABRICKS_HOST}/ai-gateway/mlflow/v1","name":"Databricks","doc":"https://docs.databricks.com/aws/en/machine-learning/foundation-models/","models":{"databricks-gpt-5-6-terra":{"id":"databricks-gpt-5-6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"databricks-claude-sonnet-4":{"id":"databricks-claude-sonnet-4","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"databricks-claude-sonnet-4-5":{"id":"databricks-claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"databricks-gemini-3-5-flash":{"id":"databricks-gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":2.7,"output":16.2,"cache_read":0.27,"input_audio":2.7},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"databricks-kimi-k2-7-code":{"id":"databricks-kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"databricks-claude-haiku-4-5":{"id":"databricks-claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"databricks-gpt-5-1":{"id":"databricks-gpt-5-1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"databricks-gemini-3-pro":{"id":"databricks-gemini-3-pro","name":"Gemini 3 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3-pro-preview"},"databricks-claude-sonnet-4-6":{"id":"databricks-claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"databricks-gpt-5-6-luna":{"id":"databricks-gpt-5-6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"databricks-gpt-5-4-mini":{"id":"databricks-gpt-5-4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"databricks-gemini-2-5-flash":{"id":"databricks-gemini-2-5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"databricks-gpt-5-5":{"id":"databricks-gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"databricks-gpt-oss-20b":{"id":"databricks-gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"databricks-gemini-3-1-flash-lite":{"id":"databricks-gemini-3-1-flash-lite","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":2.7,"cache_read":0.045,"input_audio":0.9},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"databricks-gpt-5-mini":{"id":"databricks-gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"databricks-claude-opus-4-6":{"id":"databricks-claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"databricks-gpt-5-6-sol":{"id":"databricks-gpt-5-6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"databricks-gpt-5-4-nano":{"id":"databricks-gpt-5-4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"databricks-glm-5-2":{"id":"databricks-glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"experimental":{"modes":{"fast":{"cost":{"input":2.45,"output":7.7,"cache_read":0.455},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"databricks-gpt-5-2":{"id":"databricks-gpt-5-2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"databricks-claude-opus-4-1":{"id":"databricks-claude-opus-4-1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"databricks-claude-opus-4-7":{"id":"databricks-claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"databricks-gpt-5-nano":{"id":"databricks-gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"databricks-gpt-oss-120b":{"id":"databricks-gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"databricks-claude-opus-4-5":{"id":"databricks-claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"databricks-gemini-3-1-pro":{"id":"databricks-gemini-3-1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":3.6,"output":21.6,"cache_read":0.36,"input_audio":3.6},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"databricks-gpt-5-4":{"id":"databricks-gpt-5-4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"databricks-gemini-2-5-pro":{"id":"databricks-gemini-2-5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"databricks-gpt-5":{"id":"databricks-gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"databricks-gemini-3-flash":{"id":"databricks-gemini-3-flash","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":0.9,"output":5.4,"cache_read":0.09,"input_audio":1.8},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"}}},"runinfra":{"id":"runinfra","env":["RUNINFRA_GATEWAY_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.runinfra.ai/v1","name":"RunInfra","doc":"https://runinfra.ai/docs","models":{"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.13,"output":0.27,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.6,"output":1.9,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"Inferact/Qwen3.8-2.4T-A95B-NVFP4":{"id":"Inferact/Qwen3.8-2.4T-A95B-NVFP4","name":"Qwen3.8 2.4T A95B (NVFP4)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"ornith-ai/Ornith-1.5-35B-A3B":{"id":"ornith-ai/Ornith-1.5-35B-A3B","name":"Ornith 1.5 35B A3B","description":"Mixture-of-experts coding-reasoning model for agentic software tasks, tool use, and image understanding","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-18","last_updated":"2026-08-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"deepreinforce/ornith-1.5-35b-a3b"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.15,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":1.3,"output":4.2,"cache_read":0.13},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.11,"output":0.45,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"tensorx":{"id":"tensorx","env":["TENSORX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tensorx.ai/v1","name":"TensorX","doc":"https://docs.tensorx.ai/","models":{"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"deprecated","cost":{"input":1.75,"output":3.5,"cache_read":0.4375,"cache_write":2.185},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":2,"output":4,"cache_read":0.5},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.25,"output":0.3,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1-0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":8192},"status":"deprecated","cost":{"input":0.66,"output":2.6,"cache_read":0.165,"cache_write":0.825},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"status":"deprecated","cost":{"input":0.3,"output":0.5,"cache_read":0.075,"cache_write":0.375},"canonical_model_id":"deepseek/deepseek-v3.2"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.075,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":2,"cache_read":0.1},"canonical_model_id":"minimax/MiniMax-M3"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.75},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":2.8,"cache_read":0.125,"cache_write":0.625},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.25,"output":4.5,"cache_read":0.3125},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":1,"output":4,"cache_read":0.25,"cache_write":1.25},"canonical_model_id":"moonshotai/kimi-k2.6"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":2.4,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3-235b-a22b-2507":{"id":"qwen/qwen3-235b-a22b-2507","name":"Qwen3 235B-A22B-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":262144},"status":"deprecated","cost":{"input":0.072,"output":0.464,"cache_read":0.018,"cache_write":0.09},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":0.2,"cache_read":0.0375,"cache_write":0.1875},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":3.5,"cache_read":0.125,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":2.5,"output":6,"cache_read":0.63},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.8-flash-next":{"id":"qwen/qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"status":"deprecated","cost":{"input":1,"output":3.2,"cache_read":0.25,"cache_write":1.25},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4,"cache_read":0.3,"cache_write":1.5},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":1.75,"output":4.5,"cache_read":0.44},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.35,"cache_write":1.75},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.5,"output":4.5,"cache_read":0.375},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4,"cache_read":0.3,"cache_write":1.5},"canonical_model_id":"zhipuai/glm-5v-turbo"}}},"tempr":{"id":"tempr","env":["TEMPR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.temprhq.io/v1","name":"Tempr Gateway","doc":"https://temprhq.io/docs/gateway-reference.html","models":{"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"reasoning":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-flash":{"id":"deepseek/deepseek-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"cerebras/qwen-3.8-27b":{"id":"cerebras/qwen-3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.99,"output":1.49},"canonical_model_id":"alibaba/qwen3.8-27b"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5.3-flashx":{"id":"zai/glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.7-flash":{"id":"zai/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-4.6v-flash":{"id":"zai/glm-4.6v-flash","name":"GLM-4.6V-Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-5v-turbo":{"id":"zai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5v-turbo"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.6v":{"id":"zai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai/glm-4.5-flash":{"id":"zai/glm-4.5-flash","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"groq/allam-2-7b":{"id":"groq/allam-2-7b","name":"ALLaM-2-7b","description":"ALLaM-2-7b instruction tuned model by SDAIA","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-01-23","last_updated":"2025-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"sdaia/allam-2-7b"},"groq/openai/gpt-oss-20b":{"id":"groq/openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"groq/openai/gpt-oss-120b":{"id":"groq/openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"groq/qwen/qwen3.8-27b":{"id":"groq/qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131042,"output":16384},"cost":{"input":0.8,"output":4,"cache_read":0.4},"canonical_model_id":"alibaba/qwen3.8-27b"},"minimax/MiniMax-M2.7":{"id":"minimax/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/MiniMax-M3":{"id":"minimax/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/MiniMax-M2.7-highspeed":{"id":"minimax/MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-flash-latest"},"google/gemini-embedding-001":{"id":"google/gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"cost":{"input":0.15,"output":0},"canonical_model_id":"google/gemini-embedding-001"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-flash-lite-latest"},"google/gemini-embedding-2":{"id":"google/gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1},"cost":{"input":0.2,"output":0},"canonical_model_id":"google/gemini-embedding-2"},"mistral/mistral-medium-latest":{"id":"mistral/mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-latest"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/voxtral-small-latest":{"id":"mistral/voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"mistral/voxtral-small-latest"},"mistral/mistral-medium-2604":{"id":"mistral/mistral-medium-2604","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-2604"},"mistral/mistral-embed":{"id":"mistral/mistral-embed","type":"embedding","name":"Mistral Embed","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":3072},"cost":{"input":0.1,"output":0},"canonical_model_id":"mistral/mistral-embed"},"mistral/mistral-small-latest":{"id":"mistral/mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-latest"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/mistral-small-2603":{"id":"mistral/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistral/zai-glm-5-3":{"id":"mistral/zai-glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"mistral/zai-glm-5-2":{"id":"mistral/zai-glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"xiaomi-mimo/mimo-v2.6-flash":{"id":"xiaomi-mimo/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi-mimo/mimo-v2.6-pro":{"id":"xiaomi-mimo/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi-mimo/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi-mimo/mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"anthropic/claude-haiku-4-5-20251001":{"id":"anthropic/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4-5-20250929":{"id":"anthropic/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"}}},"nova":{"id":"nova","env":["NOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.nova.amazon.com/v1","name":"Nova","doc":"https://nova.amazon.com/dev/documentation","models":{"nova-2-pro-v1":{"id":"nova-2-pro-v1","name":"Nova 2 Pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-03","last_updated":"2026-01-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"reasoning":0}},"nova-2-lite-v1":{"id":"nova-2-lite-v1","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"reasoning":0},"canonical_model_id":"amazon/nova-2-lite-v1"}}},"kenari":{"id":"kenari","env":["KENARI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://kenari.id/v1","name":"Kenari","doc":"https://kenari.id/docs","models":{"grok-imagine-image-2-0":{"id":"grok-imagine-image-2-0","type":"image","name":"Grok Imagine Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":64000,"output":0},"canonical_model_id":"xai/grok-imagine-image-2.0"},"gemini-3-7-flash":{"id":"gemini-3-7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.34285714,"output":1.71428571,"cache_read":0.03428571,"cache_write":0.01714286},"canonical_model_id":"google/gemini-3.7-flash"},"glm-5-3-flash":{"id":"glm-5-3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11428571,"output":0.34285714,"cache_read":0.02285714},"canonical_model_id":"zhipuai/glm-5.3-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.85714286,"output":8.57142857,"cache_read":0.11428571},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57142857,"output":2.85714286,"cache_read":0.28571429},"canonical_model_id":"moonshotai/kimi-k2.6"},"nano-banana-2-lite":{"id":"nano-banana-2-lite","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.57142857,"output":1.14285714,"cache_read":0.00571429},"canonical_model_id":"deepseek/deepseek-v4-pro"},"nemotron-3-nano-30b-a3b":{"id":"nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"mistral-medium-3-5:free":{"id":"mistral-medium-3-5:free","name":"Mistral Medium 3.5 (Free)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-medium-2604"},"gemini-3-1-flash-tts":{"id":"gemini-3-1-flash-tts","type":"speech","name":"Gemini 3.1 Flash TTS Preview","description":"Low-latency speech generation with steerable prompts and expressive audio tags","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-15","last_updated":"2026-04-15","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"provider":{"npm":"@ai-sdk/openai"},"canonical_model_id":"google/gemini-3.1-flash-tts-preview"},"glm-4-7-flash:free":{"id":"glm-4-7-flash:free","name":"GLM-4.7-Flash (Free)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"gemini-3-1-pro":{"id":"gemini-3-1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.2,"output":7.14285714,"cache_read":0.12,"cache_write":0.22285714},"canonical_model_id":"google/gemini-3.1-pro-preview"},"nemotron-3-super-120b-a12b:free":{"id":"nemotron-3-super-120b-a12b:free","name":"Nemotron 3 Super 120B A12B (Free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"glm-5-2":{"id":"glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.28571429,"output":3.42857143,"cache_read":0.05314286},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4-flash:free":{"id":"deepseek-v4-flash:free","name":"DeepSeek V4 Flash (Free)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.036,"output":0.15428571,"cache_read":0.00371429},"canonical_model_id":"openai/gpt-oss-20b"},"mistral-large:free":{"id":"mistral-large:free","name":"Mistral Large (Free)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-large-latest"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.15428571,"output":0.62857143,"cache_read":0.01542857},"canonical_model_id":"tencent/hy3"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":272000,"output":16384},"canonical_model_id":"openai/gpt-image-2"},"grok-imagine-image":{"id":"grok-imagine-image","type":"image","name":"Grok Imagine Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"canonical_model_id":"xai/grok-imagine-image"},"gemini-3-6-flash":{"id":"gemini-3-6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.34285714,"output":1.71428571,"cache_read":0.03428571,"cache_write":0.01714286},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-5-4-mini":{"id":"gpt-5-4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.85714286,"output":5.37142857},"canonical_model_id":"openai/gpt-5.4-mini"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15714286,"output":0.31428571,"cache_read":0.00371429},"canonical_model_id":"deepseek/deepseek-v4-flash"},"nemotron-3-ultra-550b-a55b":{"id":"nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.68571429,"output":4.28571429,"cache_read":0.06857143},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"kimi-k2-7-code":{"id":"kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57142857,"output":2.85714286,"cache_read":0.11428571},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":12,"output":57.14285714,"cache_read":1.2,"cache_write":14.85714286},"canonical_model_id":"anthropic/claude-fable-5"},"mimo-v2-5:free":{"id":"mimo-v2-5:free","name":"MiMo-V2.5 (Free)","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"minimax-m2-7-highspeed":{"id":"minimax-m2-7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.59428571,"output":2.4,"cache_read":0.11428571},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.71428571,"output":29.71428571,"cache_read":0.57142857,"cache_write":7.42857143},"canonical_model_id":"anthropic/claude-opus-4-7"},"kimi-k2-6:free":{"id":"kimi-k2-6:free","name":"Kimi K2.6 (Free)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15714286,"output":0.31428571,"cache_read":0.00371429},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gpt-5-5":{"id":"gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5.71428571,"output":36,"cache_read":0.57142857},"canonical_model_id":"openai/gpt-5.5"},"mimo-v2-5-pro":{"id":"mimo-v2-5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.52,"output":1.02857143},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.36,"output":1.42857143,"cache_read":0.06857143},"canonical_model_id":"minimax/MiniMax-M2.7"},"gpt-5-6-terra":{"id":"gpt-5-6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"status":"deprecated","cost":{"input":2.4,"output":14.28571429},"canonical_model_id":"openai/gpt-5.6-terra"},"hy3:free":{"id":"hy3:free","name":"Hy3 (Free)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"tencent/hy3"},"step-3-7-flash:free":{"id":"step-3-7-flash:free","name":"Step 3.7 Flash (Free)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"stepfun/step-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.36,"output":1.42857143,"cache_read":0.06857143},"canonical_model_id":"minimax/MiniMax-M3"},"whisper-large-v3-turbo":{"id":"whisper-large-v3-turbo","type":"transcription","name":"Whisper Large v3 Turbo","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"provider":{"npm":"@ai-sdk/openai"},"canonical_model_id":"openai/whisper-large-v3-turbo"},"kimi-k2-7-code:free":{"id":"kimi-k2-7-code:free","name":"Kimi K2.7 Code (Free)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.71428571,"output":28.57142857,"cache_read":2.28571429,"cache_write":5.71428571},"canonical_model_id":"anthropic/claude-opus-5"},"nano-banana-pro":{"id":"nano-banana-pro","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"canonical_model_id":"google/gemini-3-pro-image"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.71428571,"output":29.71428571,"cache_read":0.57142857,"cache_write":7.42857143},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.42857143,"output":7.14285714,"cache_read":0.57142857,"cache_write":1.71428571},"canonical_model_id":"anthropic/claude-sonnet-5"},"gemini-3-1-flash-lite":{"id":"gemini-3-1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.28571429,"output":1.71428571,"cache_read":0.02857143,"cache_write":0.08571429},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.036,"output":0.2,"cache_read":0.0036},"canonical_model_id":"openai/gpt-oss-120b"},"gpt-5-6-luna":{"id":"gpt-5-6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.2,"output":1.14285714,"cache_read":0.02,"cache_write":0.25714286},"canonical_model_id":"openai/gpt-5.6-luna"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.10285714,"output":0.40571429,"cache_read":0.04114286},"canonical_model_id":"google/gemma-4-31b-it"},"glm-5-1":{"id":"glm-5-1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"status":"deprecated","canonical_model_id":"zhipuai/glm-5.1"},"nemotron-3-super-120b-a12b":{"id":"nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.09714286,"output":0.48,"cache_read":0.00971429},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.28571429,"output":6.85714286,"cache_read":0.34285714},"canonical_model_id":"xai/grok-4.6"},"grok-build-0-1":{"id":"grok-build-0-1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"status":"deprecated","canonical_model_id":"xai/grok-build-0.1"},"qwen3-8-max":{"id":"qwen3-8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.4,"output":6.85714286,"cache_read":0.29714286,"cache_write":2.97142857},"canonical_model_id":"alibaba/qwen3.8-max"},"grok-4-5":{"id":"grok-4-5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.28571429,"output":6.85714286,"cache_read":0.34285714},"canonical_model_id":"xai/grok-4.5"},"gemini-2-5-flash-lite":{"id":"gemini-2-5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash-lite"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.6,"output":17.71428571,"cache_read":0.36,"cache_write":4.45714286},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-2-5-flash":{"id":"gemini-2-5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash"},"qwen3-7-plus":{"id":"qwen3-7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.38285714,"output":1.48571429,"cache_read":0.07428571,"cache_write":0.48},"canonical_model_id":"alibaba/qwen3.7-plus"},"step-3-7-flash":{"id":"step-3-7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.24,"output":1.37142857,"cache_read":0.048},"canonical_model_id":"stepfun/step-3.7-flash"},"glm-5-3":{"id":"glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.28571429,"output":3.42857143,"cache_read":0.05314286},"canonical_model_id":"zhipuai/glm-5.3"},"gemini-3-5-flash":{"id":"gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-3.5-flash"},"gpt-5-6-sol":{"id":"gpt-5-6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.97142857,"output":17.71428571,"cache_read":1.18857143,"cache_write":3.71428571},"canonical_model_id":"openai/gpt-5.6-sol"},"mimo-v2-5":{"id":"mimo-v2-5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.16571429,"output":0.33142857},"canonical_model_id":"xiaomi/mimo-v2.5"},"nano-banana-2":{"id":"nano-banana-2","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"canonical_model_id":"google/gemini-3.1-flash-image"}}},"freemodel":{"id":"freemodel","env":["FREEMODEL_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://cc.freemodel.dev/v1","name":"FreeModel","doc":"https://freemodel.dev","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175,"cache_write":1.75},"canonical_model_id":"openai/gpt-5.3-codex"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"openai/gpt-5.4-mini"}}},"bee":{"id":"bee","env":["BEE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.bee.heossi.com/bee","name":"Bee by HEOSSI","doc":"https://bee.heossi.com/docs/sdks","models":{"bee-cell":{"id":"bee-cell","name":"Bee Cell 1.0","description":"Fast hosted AI for chat, coding help, document question answering, image understanding, and lightweight technical reasoning.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.36}},"bee-swarm":{"id":"bee-swarm","name":"Bee Swarm 1.0","description":"Premium reasoning, multi-model cloud execution, large-document analysis, and high-value technical work. Includes the evolution engine and stronger eval gates by default.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":false,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":25,"output":150,"cache_read":2.5,"cache_write":30}},"bee-hive":{"id":"bee-hive","name":"Bee Hive 1.0","description":"Scalable Bee serving with quantum-reasoning routing and multi-base domain-adapter routing for shared knowledge workflows, document intelligence, and lower-cost cloud inference. Domain adapters activate per customer domain after evaluation.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":5,"output":20,"cache_read":0.5,"cache_write":6}},"bee-buzz":{"id":"bee-buzz","name":"Bee Buzz 1.0","description":"Mid-tier production model for tool-using agents, multi-step workflows, and builder/integration work. Bigger context than Comb, deeper planning, faster than Hive on routine tasks.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":24576},"cost":{"input":2.5,"output":10,"cache_read":0.25,"cache_write":3}},"bee-brood":{"id":"bee-brood","name":"Bee Brood 1.0","description":"Production multimodal reasoning for image and video analysis, tool-enabled workflows, and multi-step planning at controlled cost.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":16384},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.72}},"bee-comb":{"id":"bee-comb","name":"Bee Comb 2.0","description":"Stronger local/workstation AI for coding, architecture, security reasoning, larger RAG, and domain-specific workflows. The default workhorse for engineers and dev teams.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":16384},"cost":{"input":1.25,"output":5,"cache_read":0.125,"cache_write":1.5}}}},"auriko":{"id":"auriko","env":["AURIKO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.auriko.ai/v1","name":"Auriko","doc":"https://docs.auriko.ai","models":{"qwen-3.6-plus":{"id":"qwen-3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.1,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"minimax-m2-7-highspeed":{"id":"minimax-m2-7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"}}},"meganova":{"id":"meganova","env":["MEGANOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.meganova.ai/v1","name":"Meganova","doc":"https://docs.meganova.ai","models":{"Qwen/Qwen2.5-VL-32B-Instruct":{"id":"Qwen/Qwen2.5-VL-32B-Instruct","name":"Qwen2.5 VL 32B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.2,"output":0.6}},"Qwen/Qwen3.5-Plus":{"id":"Qwen/Qwen3.5-Plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"reasoning":2.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.09,"output":0.1},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.15},"canonical_model_id":"minimax/MiniMax-M2.5"},"MiniMaxAI/MiniMax-M2.1":{"id":"MiniMaxAI/MiniMax-M2.1","name":"MiniMax M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.28,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.19,"output":0.79},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":64000},"cost":{"input":0.5,"output":2.15,"cache_read":0.4},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-V3.2-Exp":{"id":"deepseek-ai/DeepSeek-V3.2-Exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-10","last_updated":"2025-10-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.26,"output":0.38,"cache_read":0.13}},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"deepseek/deepseek-v3-0324"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2026-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/Kimi-K2-Thinking":{"id":"moonshotai/Kimi-K2-Thinking","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":2.6},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"mistralai/Mistral-Small-3.2-24B-Instruct-2506":{"id":"mistralai/Mistral-Small-3.2-24B-Instruct-2506","name":"Mistral Small 3.2 24B Instruct","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-small-2506"},"mistralai/Mistral-Nemo-Instruct-2407":{"id":"mistralai/Mistral-Nemo-Instruct-2407","name":"Mistral Nemo Instruct 2407","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.02,"output":0.04},"canonical_model_id":"mistral/mistral-nemo"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"XiaomiMiMo/MiMo-V2-Flash":{"id":"XiaomiMiMo/MiMo-V2-Flash","name":"MiMo V2 Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32000},"status":"deprecated","cost":{"input":0.1,"output":0.3},"canonical_model_id":"xiaomi/mimo-v2-flash"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.45,"output":1.9},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.4,"output":1.75,"cache_read":0.08},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.8,"output":2.56,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"}}},"alibaba-coding-plan-cn":{"id":"alibaba-coding-plan-cn","env":["ALIBABA_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://coding.dashscope.aliyuncs.com/v1","name":"Alibaba Coding Plan (China)","doc":"https://help.aliyun.com/zh/model-studio/coding-plan","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-11","last_updated":"2026-02-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-max-2026-01-23":{"id":"qwen3-max-2026-01-23","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":24576},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}}}},"melious":{"id":"melious","env":["MELIOUS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.melious.ai/v1","name":"Melious","doc":"https://melious.ai/docs/reference/models","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"status":"deprecated","cost":{"input":1.10124,"output":3.36168,"cache_read":0.266616},"canonical_model_id":"zhipuai/glm-5"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.23184,"output":1.1592,"cache_read":0.011592},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1592,"output":3.4776,"cache_read":0.23184},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3.196275,"output":15.981375,"cache_read":0.841125},"canonical_model_id":"moonshotai/kimi-k3"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"status":"deprecated","cost":{"input":1.50696,"output":4.6368,"cache_read":0.370944},"canonical_model_id":"zhipuai/glm-5.1"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"status":"deprecated","cost":{"input":0.5796,"output":2.95596,"cache_read":0.139104},"canonical_model_id":"moonshotai/kimi-k2.5"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.728975,"output":3.644875,"cache_read":0.190655},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.33675,"output":1.347,"cache_read":0.033675},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.11592,"output":0.2898,"cache_read":0.023184},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.78505,"output":3.3645,"cache_read":0.280375},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-r1-0528":{"id":"deepseek-r1-0528","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":32768},"status":"deprecated","cost":{"input":0.69552,"output":2.78208,"cache_read":0.185472},"canonical_model_id":"deepseek/deepseek-r1"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11592,"output":0.46368,"cache_read":0.023184},"canonical_model_id":"zhipuai/glm-5.3-flash"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":64000},"status":"deprecated","cost":{"input":0.34776,"output":0.5796,"cache_read":0.092736},"canonical_model_id":"deepseek/deepseek-v3.2"}}},"zhipuai":{"id":"zhipuai","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://open.bigmodel.cn/api/paas/v4","name":"Zhipu AI","doc":"https://docs.bigmodel.cn/cn/guide/start/pricing","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.59758,"output":2.68911,"cache_read":0.1494,"cache_write":0,"tiers":[{"input":0.89637,"output":3.28669,"cache_read":0.22409,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.19516,"output":4.18306,"cache_read":0.29879,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5.3-flashx":{"id":"glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.29879,"output":1.04577,"cache_read":0.08516,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.89637,"output":3.58548,"cache_read":0.19421,"cache_write":0,"tiers":[{"input":1.19516,"output":4.18306,"cache_read":0.29879,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.29879,"output":0.89637,"cache_read":0.05976,"cache_write":0,"tiers":[{"input":0.59758,"output":1.79274,"cache_read":0.11952,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5v"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"glm-4.7-flashx":{"id":"glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.0747,"output":0.44819,"cache_read":0.01494,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.44819,"output":2.09153,"cache_read":0.08964,"cache_write":0,"tiers":[{"input":0.59758,"output":2.39032,"cache_read":0.11952,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"glm-4.6v-flash":{"id":"glm-4.6v-flash","name":"GLM-4.6V-Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.19516,"output":4.18306,"cache_read":0.29879,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11952,"output":0.41831,"cache_read":0.03436,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.74698,"output":3.28669,"cache_read":0.17927,"cache_write":0,"tiers":[{"input":1.04577,"output":3.88427,"cache_read":0.26891,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.1494,"output":0.44819,"cache_read":0.02988,"cache_write":0,"tiers":[{"input":0.29879,"output":0.89637,"cache_read":0.05976,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.11952,"output":0.89637,"cache_read":0.0239,"cache_write":0,"tiers":[{"input":0.17927,"output":1.19516,"cache_read":0.03585,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5-air"},"glm-4.5-flash":{"id":"glm-4.5-flash","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-flash"}}},"google-vertex-anthropic":{"id":"google-vertex-anthropic","env":["GOOGLE_VERTEX_PROJECT","GOOGLE_VERTEX_LOCATION","GOOGLE_APPLICATION_CREDENTIALS"],"npm":"@ai-sdk/google-vertex/anthropic","name":"Vertex (Anthropic)","doc":"https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude","models":{"claude-sonnet-4@20250514":{"id":"claude-sonnet-4@20250514","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"claude-fable-5-1@default":{"id":"claude-fable-5-1@default","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7@default":{"id":"claude-opus-4-7@default","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"claude-sonnet-5@default":{"id":"claude-sonnet-5@default","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-opus-4-5@20251101":{"id":"claude-opus-4-5@20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"claude-sonnet-4-6@default":{"id":"claude-sonnet-4-6@default","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-opus-4@20250514":{"id":"claude-opus-4@20250514","name":"Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"claude-sonnet-4-5@20250929":{"id":"claude-sonnet-4-5@20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-opus-4-8@default":{"id":"claude-opus-4-8@default","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-opus-4-6@default":{"id":"claude-opus-4-6@default","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5-5@default":{"id":"claude-opus-5-5@default","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-1@20250805":{"id":"claude-opus-4-1@20250805","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"claude-sonnet-5-5@default":{"id":"claude-sonnet-5-5@default","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5@20251001":{"id":"claude-haiku-4-5@20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"claude-haiku-5-5@default":{"id":"claude-haiku-5-5@default","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-fable-5@default":{"id":"claude-fable-5@default","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-5@default":{"id":"claude-opus-5@default","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"}}},"nearai":{"id":"nearai","env":["NEARAI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://cloud-api.near.ai/v1","name":"NEAR AI Cloud","doc":"https://docs.near.ai/","models":{"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":1.1,"output":1.1,"cache_read":0.55},"canonical_model_id":"deepseek/deepseek-v3.2"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3-VL 30B-A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":8192},"cost":{"input":0.15,"output":0.55,"cache_read":0.03},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3-Reranker-0.6B":{"id":"Qwen/Qwen3-Reranker-0.6B","type":"reranking","name":"Qwen3 Reranker 0.6B","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":1024},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"alibaba/qwen3-reranker-0.6b"},"Qwen/Qwen3.6-35B-A3B-FP8":{"id":"Qwen/Qwen3.6-35B-A3B-FP8","name":"Qwen 3.6 35B A3B FP8","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.17,"output":1.1,"cache_read":0.056},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-Embedding-0.6B":{"id":"Qwen/Qwen3-Embedding-0.6B","type":"embedding","name":"Qwen3 Embedding 0.6B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":1024},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"alibaba/qwen3-embedding-0.6b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.04,"output":1.35,"cache_read":0.018},"canonical_model_id":"alibaba/qwen3.8-27b"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"black-forest-labs/FLUX.2-klein-4B":{"id":"black-forest-labs/FLUX.2-klein-4B","type":"image","name":"FLUX.2 Klein 4B","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"canonical_model_id":"black-forest-labs/flux-2-klein-4b"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"openai/whisper-large-v3":{"id":"openai/whisper-large-v3","type":"transcription","name":"Whisper Large v3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"openai/whisper-large-v3"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3.3,"output":16.5,"cache_read":0.33},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.81,"output":3.85,"cache_read":0.41},"canonical_model_id":"moonshotai/kimi-k2.6"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":7.5},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.5,"output":3.3,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.8,"output":5.75,"cache_read":0.45},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.08,"output":0.5,"cache_read":0.012},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"the-grid-ai":{"id":"the-grid-ai","env":["THEGRID_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.thegrid.ai/v1","name":"The Grid AI","doc":"https://thegrid.ai/docs","models":{"gemini-pro-latest":{"id":"gemini-pro-latest","name":"Gemini Pro Latest","description":"Lab latest market for Gemini Pro frontier supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":65536}},"text-prime":{"id":"text-prime","name":"Text Prime","description":"Reliable models for everyday text generation, editing, and analysis across diverse workflows. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-26","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"input":120000,"output":131072}},"bytedance-pro-latest":{"id":"bytedance-pro-latest","name":"ByteDance Pro Latest","description":"Lab latest market for ByteDance Pro supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":256000,"output":131072}},"text-standard":{"id":"text-standard","name":"Text Standard","description":"Price-optimized models with low-latency, high-throughput and shorter maximum outputs. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-26","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":120000,"output":65536}},"text-max":{"id":"text-max","name":"Text Max","description":"Frontier models for deep reasoning, long context, and complex workflows. Any model that meets the contract spec can serve your request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000}},"glm-latest":{"id":"glm-latest","name":"GLM Latest","description":"Lab latest market for GLM supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":917504,"output":131072}},"agent-max":{"id":"agent-max","name":"Agent Max","description":"Frontier models for autonomous research, deep multi-step tool chains, and complex long-horizon tasks. Any model that meets the contract spec can serve your request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-05-04","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"status":"beta"},"code-prime":{"id":"code-prime","name":"Code Prime","description":"Reliable models for everyday software tasks, code completion, review, and standard debugging. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"input":120000,"output":131072},"status":"beta"},"minimax-latest":{"id":"minimax-latest","name":"MiniMax Latest","description":"Lab latest market for MiniMax supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":512000}},"agent-standard":{"id":"agent-standard","name":"Agent Standard","description":"Price-optimized models for fast tool calls, simple agent loops, high-throughput automation, and orchestration. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":120000,"output":65536},"status":"beta"},"deepseek-pro-latest":{"id":"deepseek-pro-latest","name":"DeepSeek Pro Latest","description":"Lab latest market for DeepSeek Pro supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":384000}},"kimi-latest":{"id":"kimi-latest","name":"Kimi Latest","description":"Lab latest market for Kimi supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":922000,"output":943718}},"code-max":{"id":"code-max","name":"Code Max","description":"Frontier models for complex research, architectural decisions, debugging, and multi-file development. Any model that meets the contract spec can serve your request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-05-04","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"status":"beta"},"code-standard":{"id":"code-standard","name":"Code Standard","description":"Price-optimized models for rapid autocomplete, linting, high-frequency suggestions, and batch edits. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":120000,"output":65536},"status":"beta"},"gpt-sol-latest":{"id":"gpt-sol-latest","name":"GPT Sol Latest","description":"Lab latest market for GPT frontier supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000}},"claude-opus-latest":{"id":"claude-opus-latest","name":"Claude Opus Latest","description":"Lab latest market for Claude Opus frontier supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000}},"agent-prime":{"id":"agent-prime","name":"Agent Prime","description":"Reliable models for dependable agentic applications, multi-step tool use, and reasoning workflows. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"input":120000,"output":131072},"status":"beta"}}},"cortecs":{"id":"cortecs","env":["CORTECS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.cortecs.ai/v1","name":"Cortecs","doc":"https://api.cortecs.ai/v1/models","models":{"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":500000},"cost":{"input":0.1,"output":0.4,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen2.5-vl-72b-instruct":{"id":"qwen2.5-vl-72b-instruct","name":"qwen2.5-vl-72b-instruct","description":"Qwen2.5-VL is a powerful vision-language model with advanced capabilities in visual understanding, long video reasoning, and structured output generation.","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":1.014,"output":1.014},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"gemma-3-27b-it":{"id":"gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Gemma 3 is a family of lightweight, multimodal models from Google, supporting text and image inputs, multilingual capabilities, and a 131K context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":110000},"cost":{"input":0.099,"output":0.299},"canonical_model_id":"google/gemma-3-27b-it"},"claude-haiku-5.5":{"id":"claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.113,"output":0.563,"cache_read":0.011,"cache_write":0.14},"canonical_model_id":"anthropic/claude-haiku-5-5"},"codestral-2508":{"id":"codestral-2508","name":"Codestral 25.08","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.368,"output":1.103,"cache_read":0.037},"canonical_model_id":"mistral/codestral-2508"},"kolibri-1":{"id":"kolibri-1","name":"kolibri-1","description":"Kolibri-1 is Aleph Alpha’s open-weight English-German MoE language model, designed for enterprise use with efficient inference and advanced reasoning.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-22","last_updated":"2026-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":81920},"cost":{"input":0.167,"output":0.613,"cache_read":0.084}},"qwen3guard-gen-8b":{"id":"qwen3guard-gen-8b","name":"qwen3guard-gen-8b","description":"Qwen3Guard-Gen-8B is a large-scale multilingual safety moderation model designed for high-accuracy prompt and response classification.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0,"output":0}},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.272,"output":1.631,"cache_read":0.025,"cache_write":0.082},"canonical_model_id":"google/gemini-3.1-flash-lite"},"minicpm-v-4.5":{"id":"minicpm-v-4.5","name":"minicpm-v-4.5","description":"MiniCPM-V 4.5 is a compact, high-performance vision-language model excelling in video understanding, OCR, and multimodal reasoning with efficient deployment.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.651,"output":1.097}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.201,"output":0.6,"cache_read":0.004},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.511,"output":2.045,"cache_read":0.154},"canonical_model_id":"zhipuai/glm-5.3"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.724,"output":0.724},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"mixtral-8x7B-instruct-v0.1":{"id":"mixtral-8x7B-instruct-v0.1","name":"Mixtral 8x7B Instruct v0.1","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mistral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.488,"output":0.758},"canonical_model_id":"mistral/mixtral-8x7b-instruct-v0.1"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.199,"cache_read":0.219,"cache_write":2.749},"canonical_model_id":"openai/gpt-5.6-terra"},"nova-2-lite":{"id":"nova-2-lite","name":"Nova 2 Lite","description":"Nova 2 Lite is an advanced multimodal reasoning model that combines efficiency and performance, delivering reliable AI for agentic workflows and enterprise applications.","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.373,"output":3.144},"canonical_model_id":"amazon/nova-2-lite"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":14.999},"canonical_model_id":"moonshotai/kimi-k3"},"mistral-nemo-instruct-2407":{"id":"mistral-nemo-instruct-2407","name":"mistral-nemo-instruct-2407","description":"A 12B parameter, instruct-tuned language model by Mistral AI and NVIDIA, designed for advanced instruction following, multi-turn conversations, and generating text and code across multiple languages.","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":118000,"output":118000},"cost":{"input":0.145,"output":0.145,"cache_read":0.014},"canonical_model_id":"mistral/mistral-nemo"},"gpt-oss-safeguard-120b":{"id":"gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.179,"output":0.697},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"output":196000},"cost":{"input":0.348,"output":1.39,"cache_read":0.075},"canonical_model_id":"minimax/MiniMax-M2.5"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.7,"output":3.4,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.179,"output":0.697,"cache_read":0.014},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"cost":{"input":1.384,"output":4.348,"cache_read":0.346},"canonical_model_id":"zhipuai/glm-5.1"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.179,"output":0.697},"canonical_model_id":"alibaba/qwen3-32b"},"claude-opus4-7":{"id":"claude-opus4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.437,"output":27.186,"cache_read":0.544,"cache_write":6.797},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":11.999,"cache_read":0.12,"cache_write":3},"canonical_model_id":"openai/gpt-6.1-sol"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.495,"output":9.964,"cache_read":0.242,"cache_write":0.434},"canonical_model_id":"google/gemini-2.5-pro"},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"qwen3-vl-235b-a22b","description":"Qwen3 VL 235B A22B is a 235B-parameter MoE vision-language flagship model (≈22B active) designed for frontier-level multimodal understanding across text, images, documents, and long videos.","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.617,"output":3.119,"cache_read":0.052},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0.996,"output":4.982,"cache_read":0.099,"cache_write":1.186},"canonical_model_id":"anthropic/claude-haiku-4-5"},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.613,"output":1.838,"cache_read":0.061},"canonical_model_id":"mistral/mistral-large-2512"},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Ministral 3 14B is a frontier-level 14B multimodal model optimized for local deployment, delivering state-of-the-art text and vision reasoning with a 256K context window and strong agentic capabilities.","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.24,"output":0.24,"cache_read":0.022},"canonical_model_id":"mistral/ministral-14b-2512"},"devstral-2512":{"id":"devstral-2512","name":"Devstral 2","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.478,"output":2.392,"cache_read":0.045},"canonical_model_id":"mistral/devstral-2512"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.045,"output":0.167},"canonical_model_id":"openai/gpt-oss-20b"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.12,"output":0.6,"cache_read":0.012,"cache_write":0.15},"canonical_model_id":"openai/gpt-6-luna"},"apertus-70b":{"id":"apertus-70b","name":"Apertus 70B","description":"Apertus 70B is an open, multilingual language model designed for research, long-context reasoning, and sovereignty-focused AI systems.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":1.393,"output":2.228},"canonical_model_id":"swiss-ai/apertus-70b"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":203000},"cost":{"input":0.08,"output":0.478},"canonical_model_id":"zhipuai/glm-4.7-flash"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.111,"output":0.434,"cache_read":0.056},"canonical_model_id":"openai/gpt-4.1-nano"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.199,"output":0.598,"cache_read":0.018},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.898,"output":15.453,"cache_read":0.242},"canonical_model_id":"openai/gpt-5.4"},"nova-lite-v1":{"id":"nova-lite-v1","name":"Nova Lite 1.0","description":"Nova Lite is a fast, low-cost multimodal foundation model capable of reasoning over text, images, and video in 200+ languages.","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.069,"output":0.275},"canonical_model_id":"amazon/nova-lite-v1"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.219,"output":1.32,"cache_read":0.022,"cache_write":0.275},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.038},"canonical_model_id":"google/gemini-3.6-flash"},"ministral-3b-2512":{"id":"ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Ministral 3 3B is a compact, efficient multimodal model with strong language, vision capabilities, and ideal for custom fine-tuning.","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.123,"output":0.123,"cache_read":0.012},"canonical_model_id":"mistral/ministral-3b-2512"},"claude-opus4-6":{"id":"claude-opus4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.313,"output":26.561,"cache_read":0.531,"cache_write":6.645},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5.5":{"id":"claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.399,"output":21.998,"cache_read":0.219,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.656,"output":3.3,"cache_read":0.18},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"nvidia-nemotron-3-nano-30b-a3b":{"id":"nvidia-nemotron-3-nano-30b-a3b","name":"nvidia-nemotron-3-nano-30b-a3b","description":"Nemotron-Nano-3-30B-A3B is a compact Mixture-of-Experts model optimized for efficient reasoning, chat, and coding, with strong multilingual support and long-context RAG and agent workflows.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":81920},"cost":{"input":0.111,"output":0.557},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.33,"output":2.749,"cache_read":0.033,"cache_write":0.082},"canonical_model_id":"google/gemini-3.5-flash-lite"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.111,"output":0.167},"canonical_model_id":"alibaba/qwen3.5-9b"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.159,"output":0.638,"cache_read":0.081},"canonical_model_id":"openai/gpt-4o-mini"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.279,"output":1.114,"cache_read":0.111},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":2.25,"output":11.249,"cache_read":0.113},"canonical_model_id":"google/gemini-4-argon"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":2,"output":3.999,"cache_read":0.5},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"minimax-m2.1":{"id":"minimax-m2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"output":196000},"cost":{"input":0.359,"output":1.435},"canonical_model_id":"minimax/MiniMax-M2.1"},"claude-sonnet-5.5":{"id":"claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.25,"output":11.249,"cache_read":0.225,"cache_write":2.813},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.434,"output":1.704,"cache_read":0.134},"canonical_model_id":"openai/gpt-4.1-mini"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.279,"output":2.192,"cache_read":0.056},"canonical_model_id":"openai/gpt-5-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":12.272,"output":61.359,"cache_read":1.228,"cache_write":15.34},"canonical_model_id":"openai/gpt-6-astra"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"mistral-small-3.2-24b-instruct-2506","description":"Mistral-Small-3.2-24B-Instruct-2506 is a 24B parameter instruction-tuned model with enhanced long-context support (128k) and state-of-the-art vision understanding.","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.1,"output":0.312},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistral-large-4-0":{"id":"mistral-large-4-0","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":0.765,"output":2.351,"cache_read":0.079},"canonical_model_id":"mistral/mistral-large-4"},"qwen3-30b-a3b-instruct-2507":{"id":"qwen3-30b-a3b-instruct-2507","name":"qwen3-30b-a3b-instruct-2507","description":"Qwen3-30B-A3B-Instruct-2507 is an advanced Mixture-of-Experts model optimized for reasoning, coding, and multilingual instruction following.","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.099,"output":0.299},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.668,"output":4.01},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.082,"output":0.413,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"mistral-medium-3.5":{"id":"mistral-medium-3.5","name":"mistral-medium-3.5","description":"Mistral Medium 3.5 is a frontier multimodal 128B model combining reasoning, coding, and instruction-following with strong agentic performance and efficient deployment.","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1.532,"output":7.843,"cache_read":0.154},"canonical_model_id":"mistral/mistral-medium-3-5"},"claude-opus4-5":{"id":"claude-opus4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.313,"output":26.568,"cache_read":0.531,"cache_write":6.645},"canonical_model_id":"anthropic/claude-opus-4-5"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.475,"output":2.97,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.6"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.038},"canonical_model_id":"google/gemini-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":10.96,"cache_read":0.156},"canonical_model_id":"openai/gpt-5.1"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2.5,"output":6,"cache_read":0.625},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.498,"cache_read":0.55,"cache_write":6.874},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus4-8":{"id":"claude-opus4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.437,"output":27.186,"cache_read":0.544,"cache_write":6.797},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-4-5-sonnet":{"id":"claude-4-5-sonnet","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":2.989,"output":14.945,"cache_read":0.326,"cache_write":4.078},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.446,"output":3.008},"canonical_model_id":"alibaba/qwen3.6-27b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1000000},"cost":{"input":0.9,"output":3.24,"cache_read":0.189},"canonical_model_id":"zhipuai/glm-5.2"},"nova-micro-v1":{"id":"nova-micro-v1","name":"Nova Micro 1.0","description":"Nova Micro is a multilingual text-to-text foundation model with strong reasoning capabilities and broad language coverage across 200+ languages.","family":"nova-micro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.04,"output":0.159},"canonical_model_id":"amazon/nova-micro-v1"},"qwen3.8-flash-next":{"id":"qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.201,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"mistral-small-2503":{"id":"mistral-small-2503","name":"mistral-small-2503","description":"Combines advanced text and vision capabilities with 24 billion parameters, supporting multilingual tasks and long contexts up to 131k tokens, making it versatile for various applications without sacrificing performance.","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.111,"output":0.334},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":2.659,"output":10.635,"cache_read":1.33},"canonical_model_id":"openai/gpt-4o"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.35,"cache_read":0.018},"canonical_model_id":"zhipuai/glm-5.3-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.219,"cache_write":2.749},"canonical_model_id":"anthropic/claude-sonnet-5"},"mistral-large-2402":{"id":"mistral-large-2402","name":"mistral-large-2402","description":"Mistral Large (24.02) is Mistral AI’s most advanced language model, built for complex multilingual reasoning, code generation, and deep text understanding.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-05-26","last_updated":"2025-05-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":4.284,"output":12.952}},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.299,"output":2.491,"cache_read":0.029,"cache_write":0.097},"canonical_model_id":"google/gemini-2.5-flash"},"voxtral-small-2507":{"id":"voxtral-small-2507","name":"voxtral-small-2507","description":"Voxtral Small is a multimodal model with audio input, combining advanced speech capabilities with strong text performance for transcription, translation, and audio understanding.","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.123,"output":0.368,"cache_read":0.012},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.089,"output":0.446,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-120b"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.223,"output":0.39},"canonical_model_id":"google/gemma-4-31b-it"},"hermes-4-405b":{"id":"hermes-4-405b","name":"Hermes 4 405B","description":"Hermes 4 405B is a frontier hybrid-mode reasoning model built on Llama 3.1, optimized for advanced logic, math, coding, and structured output generation.","family":"hermes","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.996,"output":2.989},"canonical_model_id":"nousresearch/hermes-4-405b"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":10.96,"cache_read":0.156},"canonical_model_id":"openai/gpt-5"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.06,"output":0.439,"cache_read":0.019},"canonical_model_id":"openai/gpt-5-nano"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.156,"output":0.625,"cache_read":0.016},"canonical_model_id":"mistral/mistral-small-2603"},"minimax-m2":{"id":"minimax-m2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":400000,"output":196000},"cost":{"input":0.349,"output":1.405},"canonical_model_id":"minimax/MiniMax-M2"},"pixtral-large-2502":{"id":"pixtral-large-2502","name":"Pixtral Large (25.02)","description":"Pixtral Large (25.02) is a 124B open-weight multimodal model built on Mistral Large 2, offering advanced image understanding and strong performance across text and code tasks.","family":"pixtral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":1.993,"output":5.978},"canonical_model_id":"mistral/pixtral-large-2502"},"qwen3guard-gen-0.6b":{"id":"qwen3guard-gen-0.6b","name":"qwen3guard-gen-0.6b","description":"Qwen3Guard-Gen-0.6B is a lightweight multilingual safety moderation model that classifies prompts and responses into safe, controversial, or unsafe categories.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0,"output":0}},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.825,"output":4.125,"cache_read":0.082,"cache_write":0.084},"canonical_model_id":"google/gemini-3.8-flash"},"nova-pro-v1":{"id":"nova-pro-v1","name":"Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.918,"output":3.671},"canonical_model_id":"amazon/nova-pro-v1"},"claude-4-6-sonnet":{"id":"claude-4-6-sonnet","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.196,"output":15.94,"cache_read":0.32,"cache_write":3.999},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.167,"output":0.891},"canonical_model_id":"alibaba/qwen3-coder-next"},"nemotron-nano-v2-12b":{"id":"nemotron-nano-v2-12b","name":"nemotron-nano-v2-12b","description":"NVIDIA Nemotron Nano v2 12B is a 12-billion-parameter multimodal reasoning model designed for advanced video understanding, document intelligence, and visual reasoning, built with a hybrid Transformer-Mamba architecture for high efficiency and low latency.","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.24,"output":0.707},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.399,"output":21.998,"cache_read":0.44,"cache_write":5.5},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.649,"output":9.899,"cache_read":0.165,"cache_write":1},"canonical_model_id":"google/gemini-3.5-flash"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":32768},"cost":{"input":0.167,"output":0.557},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":11.999,"cache_read":0.24,"cache_write":3},"canonical_model_id":"openai/gpt-6-sol"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.5,"output":1.499,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v3.2"},"ministral-8b-2512":{"id":"ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Ministral 3 8B is a balanced, efficient multimodal model offering strong text and vision capabilities, optimized for edge and local deployment.","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.179,"output":0.179,"cache_read":0.017},"canonical_model_id":"mistral/ministral-8b-2512"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2.192,"output":8.769,"cache_read":0.546},"canonical_model_id":"openai/gpt-4.1"},"mistral-7b-instruct-v0.2":{"id":"mistral-7b-instruct-v0.2","name":"mistral-7b-instruct-v0.2","description":"Mistral 7B Instruct is a compact, 7B parameter model optimized for fast and efficient text and code generation with a 32K token context window.","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-05-26","last_updated":"2025-05-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":0.159,"output":0.219}}}},"kilo":{"id":"kilo","env":["KILO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.kilo.ai/api/gateway","name":"Kilo Gateway","doc":"https://kilo.ai","models":{"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":80000},"cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the cost-efficient tier of the V4.1 family. DeepSeek reports that it exceeds V4 Pro on performance, speed, and task...","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":384000},"cost":{"input":1.6,"output":3.2,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-chat-v3-0324":{"id":"deepseek/deepseek-chat-v3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":115200},"cost":{"input":0.29,"output":1.14,"cache_read":0.11},"canonical_model_id":"deepseek/deepseek-chat-v3-0324"},"deepseek/deepseek-chat-v3.1":{"id":"deepseek/deepseek-chat-v3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-chat-v3.1"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":943717},"cost":{"input":0.091,"output":0.182,"cache_read":0.0182},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"DeepSeek V4.1 Flash 1017 is the October 17, 2026 snapshot of DeepSeek V4.1 Flash, served by DeepSeek under the `deepseek-flash` API model name. V4.1 Flash is a sparse mixture-of-experts model...","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro.","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of [DeepSeek V4 Flash 0731](https://openrouter.ai/deepseek/deepseek-v4-flash-0731) from DeepSeek, adding image understanding while matching the base model on text capabilities including agents,...","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.00686},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-v3.1-terminus":{"id":"deepseek/deepseek-v3.1-terminus","name":"DeepSeek: DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":1}},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows.","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.2574,"output":1.0287},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek: DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":0.41}},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v3.2"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Fugu Max is the cost-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to route...","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/sakana-namazu":{"id":"sakana/sakana-namazu","name":"Sakana Namazu","description":"Sakana Namazu is a Japanese-specialized reasoning model from Sakana AI, based on Kimi K2.6 with additional training for Japanese language and business contexts. It is suited for Japanese instruction following,...","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Fugu Ultra v2 is the higher-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to...","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Fugu Ultra is the higher-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to route...","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"inception/mercury-3":{"id":"inception/mercury-3","name":"Inception: Mercury 3","description":"Mercury 3 is a diffusion-based language model from Inception, and the lab's most capable model to date. It is suited for agentic workflows and real-time applications where latency compounds, such...","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2,"output":0.7,"cache_read":0.02}},"inception/mercury-2":{"id":"inception/mercury-2","name":"Inception: Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Inception: Mercury 2.5","description":"Mercury 2.5 is the fastest reasoning LLM, and the latest diffusion LLM (dLLM) from Inception. Instead of generating tokens sequentially, Mercury 2.5 produces and refines multiple tokens in parallel, achieving...","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.2,"output":0.75,"cache_read":0.02},"canonical_model_id":"inception/mercury-2.5"},"scaledown/summarize":{"id":"scaledown/summarize","name":"ScaleDown: Summarize","description":"Summarize is an abstractive summarization model from ScaleDown. It condenses long text and documents into a fluent, faithful rewrite rather than lifting sentences from the source, and is suited for...","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":4000},"cost":{"input":0.05,"output":0}},"scaledown/extract":{"id":"scaledown/extract","name":"ScaleDown: Extract","description":"Extract is a description-driven entity extraction model from ScaleDown. Instead of a fixed set of entity types, callers describe the fields they want in plain English through `provider.options.scaledown.entities`, and the...","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":0.05,"output":0}},"~x-ai/grok-latest":{"id":"~x-ai/grok-latest","name":"xAI: Grok Latest","description":"This model always redirects to the latest Grok model from xAI.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5}},"aion-labs/aion-3.5":{"id":"aion-labs/aion-3.5","name":"AionLabs: Aion 3.5","description":"Aion 3.5 is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It uses a collaborative generation process in which multiple specialized models each...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75}},"aion-labs/aion-2.0":{"id":"aion-labs/aion-2.0","name":"AionLabs: Aion-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-21","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.8,"output":1.6,"cache_read":0.2},"canonical_model_id":"aion-labs/aion-2.0"},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"AionLabs: Aion-RP 1.0 (8B)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.8,"output":1.6},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"aion-labs/aion-3.0-mini":{"id":"aion-labs/aion-3.0-mini","name":"AionLabs: Aion-3.0-Mini","description":"Aion-3.0 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the DeepSeek family of models. It uses a collaborative generation process in which multiple specialized models each...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18},"canonical_model_id":"aion-labs/aion-3.0-mini"},"aion-labs/aion-3.0":{"id":"aion-labs/aion-3.0","name":"AionLabs: Aion-3.0","description":"Aion-3.0 is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It uses a collaborative generation process in which multiple specialized models each contribute...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75},"canonical_model_id":"aion-labs/aion-3.0"},"aion-labs/aion-3.5-mini":{"id":"aion-labs/aion-3.5-mini","name":"AionLabs: Aion 3.5 Mini","description":"Aion 3.5 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It is the smaller, lower-cost sibling of Aion 3.5 and uses...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18}},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"Morph: Morph V3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.8,"output":1.2},"canonical_model_id":"morph/morph-v3-fast"},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"Morph: Morph V3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.9,"output":1.9},"canonical_model_id":"morph/morph-v3-large"},"moondream/moondream3-preview":{"id":"moondream/moondream3-preview","name":"Moondream: Moondream 3 Preview","description":"Moondream 3 Preview is a vision-language model from Moondream with 9B total parameters and 2B active (mixture-of-experts). It handles visual question answering, captioning, OCR, and visual reasoning over images, with...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"moondream/moondream3.1-9b-a2b":{"id":"moondream/moondream3.1-9b-a2b","name":"Moondream: Moondream 3.1","description":"Moondream 3.1 is a vision-language model from Moondream, the successor to Moondream 3 Preview, with 9B total parameters and 2B active (mixture-of-experts). It handles visual question answering, captioning, OCR, and...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"reflection/beam-1.0-501b-a23b":{"id":"reflection/beam-1.0-501b-a23b","name":"Reflection: Beam 1.0 501B A23B","description":"Beam 1.0 is Reflection's open-weight reasoning model for coding and long-horizon agentic workflows. It uses a sparse mixture-of-experts architecture with 501B total and 23B active parameters per token, and a...","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.24,"output":0.98,"cache_read":0.035},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"reflection/beam-free":{"id":"reflection/beam-free","name":"Reflection: Beam 1.0 (free)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. Beam is Reflection's reasoning model for coding and agentic workflows, combining strong coding capabilities with fast, low-latency inference, served through Vercel AI Gateway at no cost. It supports function calling, structured outputs, and streaming, and takes text input only. Zero Data Retention is not available for this model.","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"reasoning":0,"cache_read":0}},"cognitivecomputations/dolphin-mistral-24b-venice-edition":{"id":"cognitivecomputations/dolphin-mistral-24b-venice-edition","name":"Venice: Uncensored","description":"Venice Uncensored Dolphin Mistral 24B Venice Edition is a fine-tuned variant of Mistral-Small-24B-Instruct-2501, developed by dphn.ai in collaboration with Venice.ai. This model is designed as an “uncensored” instruct-tuned LLM, preserving...","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2,"output":0.9},"canonical_model_id":"cognitivecomputations/dolphin-mistral-24b-venice-edition"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"Grok 4.6 is SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.20-multi-agent":{"id":"x-ai/grok-4.20-multi-agent","name":"SpaceXAI: Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Grok coding model for agentic engineering, edits, and codebase workflows","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":230400},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"SpaceXAI: Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"Grok 4.7 is SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"Grok 4.5 is SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total. It is designed for general-purpose reasoning, coding, agentic and tool-use systems,...","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":471859},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of...","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"relace/relace-apply-3":{"id":"relace/relace-apply-3","name":"Relace: Relace Apply 3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.85,"output":1.25},"canonical_model_id":"relace/relace-apply-3"},"relace/relace-search":{"id":"relace/relace-search","name":"Relace: Relace Search","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":1,"output":3},"canonical_model_id":"relace/relace-search"},"~z-ai/glm-flash-latest":{"id":"~z-ai/glm-flash-latest","name":"Z.ai: GLM Flash Latest","description":"This model always redirects to the latest model in the GLM Flash family.","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.0199,"output":0.122,"cache_read":0.0199}},"~z-ai/glm-latest":{"id":"~z-ai/glm-latest","name":"Z.ai: GLM Latest","description":"This model always redirects to the latest GLM model from Z.ai.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-19","last_updated":"2026-08-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.079,"output":1.4,"cache_read":0.078}},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"inclusionAI: Ling 3.1 Flash","description":"Ling 3.1 Flash is a hybrid reasoning mixture-of-experts model from inclusionAI, with 25B active parameters out of 560B total.","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"inclusionai/ling-3.1-flash"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"inclusionAI: Ling 3.0 Flash VL","description":"Ling 3.0 Flash VL builds on Ling 3.0 Flash (124B total / 5.5B active MoE from InclusionAI), further strengthening its language capabilities while adding native visual perception and advanced visual...","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"inclusionAI: Ling 3.0 Flash","description":"*Ling-3.0-flash* is a *124B-parameter Mixture-of-Experts (MoE) model*, with approximately *5.1B parameters activated per token*. The model is designed with *token efficiency and production-scale agentic inference* as key priorities, enabling developers...","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"sao10k/l3-lunaris-8b":{"id":"sao10k/l3-lunaris-8b","name":"Sao10K: Llama 3 8B Lunaris","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-13","last_updated":"2024-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":7372},"cost":{"input":0.04,"output":0.05},"canonical_model_id":"sao10k/l3-lunaris-8b"},"sao10k/l3.1-euryale-70b":{"id":"sao10k/l3.1-euryale-70b","name":"Sao10K: Llama 3.1 Euryale 70B v2.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.85,"output":0.85},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":32768},"cost":{"input":0.27,"output":0.95,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-01":{"id":"minimax/minimax-01","name":"MiniMax: MiniMax-01","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000192,"output":900172},"cost":{"input":0.2,"output":1.1}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.224,"output":0.88,"cache_read":0.0448},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax: MiniMax M3.1 Flash","description":"MiniMax-M3.1-Flash is a multimodal model from MiniMax. Compared with M3, it delivers significant improvements across multiple benchmarks, particularly in complex agentic tasks, coding, and cybersecurity. It is built for complex,...","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":524288},"cost":{"input":0.1,"output":0.55,"cache_read":0.01},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/minimax-m1":{"id":"minimax/minimax-m1","name":"MiniMax: MiniMax M1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":40000},"cost":{"input":0.4,"output":2.2}},"minimax/minimax-m2-her":{"id":"minimax/minimax-m2-her","name":"MiniMax-M2 Her","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":2048},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2-Her"},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"IBM: Granite 4.2 8B","description":"Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning. It supports full, low-effort,...","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"ibm-granite/granite-4.0-h-micro":{"id":"ibm-granite/granite-4.0-h-micro","name":"IBM: Granite 4.0 Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"granite","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131000,"output":117900},"cost":{"input":0.017,"output":0.112}},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"output":5529},"cost":{"input":0.35,"output":0.65},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-r","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-a-plus":{"id":"cohere/command-a-plus","name":"Cohere: Command A+","description":"Command A+ is Cohere's flagship model for enterprise agentic workflows. It accepts text and image inputs with a 192K context window, supports native tool calling with strict tool schemas, structured...","family":"command-a","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04-01","release_date":"2026-05-20","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":64000},"cost":{"input":0.3,"output":1.5,"cache_read":0.15},"canonical_model_id":"cohere/command-a-plus-05-2026"},"cohere/north-mini-code:free":{"id":"cohere/north-mini-code:free","name":"Cohere: North Mini Code (free)","description":"North Mini Code is Cohere's first agentic coding model and the debut of its North family. A sparse mixture-of-experts model with 30B total parameters and 3B active, it is optimized...","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"bytedance-seed/seed-1.6":{"id":"bytedance-seed/seed-1.6","name":"ByteDance Seed: Seed 1.6 (retires Nov 11)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.25,"output":2},"canonical_model_id":"bytedance-seed/seed-1-6"},"bytedance-seed/seed-2.0-mini":{"id":"bytedance-seed/seed-2.0-mini","name":"Seed 2.0 Mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance-seed/seed-1.6-flash":{"id":"bytedance-seed/seed-1.6-flash","name":"ByteDance Seed: Seed 1.6 Flash (retires Nov 11)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.075,"output":0.3},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"bytedance-seed/seed-2-1-turbo":{"id":"bytedance-seed/seed-2-1-turbo","name":"ByteDance Seed: Seed 2.1 Turbo","description":"Seed 2.1 Turbo is a multimodal model from ByteDance Seed for coding and long-horizon agent workflows. It is suited for end-to-end software delivery, multi-step task execution, and understanding visual and...","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":2.5},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"Seed 2.0 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.25,"output":2},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"Seed 2.0 Code","description":"Seed 2.0 Code is a model from ByteDance Seed optimized for agentic coding. It is suited for frontend development, multilingual programming tasks, and coding-agent workflows in tools such as Claude...","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.5,"output":3},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"microsoft/phi-4":{"id":"microsoft/phi-4","name":"Microsoft: Phi 4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-01-10","last_updated":"2025-01-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":14745},"cost":{"input":0.056,"output":0.112}},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Poolside: Laguna S 2.1 (retires Oct 31)","description":"Laguna S 2.1 is the latest coding agent model from [Poolside](). Laguna S 2.1 is a 118B total parameter model with 8B active parameters, scoring 70.2% on Terminal-Bench 2.1 and...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.1,"output":0.2,"cache_read":0.01},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-xs-2.1:free":{"id":"poolside/laguna-xs-2.1:free","name":"Poolside: Laguna XS 2.1 (free)","description":"Laguna XS 2.1 is the latest coding agent model in the 33B-A3B category from [Poolside](https://poolside.ai/) and a step forward from their Laguna XS.2 model (released in April 2026). It combines...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Poolside: Laguna XS 2.1 (retires Oct 31)","description":"Laguna XS 2.1 is the latest coding agent model in the 33B-A3B category from [Poolside](https://poolside.ai/) and a step forward from their Laguna XS.2 model (released in April 2026). It combines...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.2,"cache_read":0.05},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-s-2.1:free":{"id":"poolside/laguna-s-2.1:free","name":"Poolside: Laguna S 2.1 (free)","description":"Laguna S 2.1 is the latest coding agent model from [Poolside](). Laguna S 2.1 is a 118B total parameter model with 8B active parameters, scoring 70.2% on Terminal-Bench 2.1 and...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-s-2.1"},"~anthropic/claude-fable-latest":{"id":"~anthropic/claude-fable-latest","name":"Anthropic: Claude Fable Latest ($$$$)","description":"This model always redirects to the latest model in the Claude Fable family.","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5}},"~anthropic/claude-sonnet-latest":{"id":"~anthropic/claude-sonnet-latest","name":"Anthropic: Claude Sonnet Latest","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"~anthropic/claude-haiku-latest":{"id":"~anthropic/claude-haiku-latest","name":"Anthropic: Claude Haiku Latest","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"~anthropic/claude-opus-latest":{"id":"~anthropic/claude-opus-latest","name":"Anthropic: Claude Opus Latest","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5}},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo-V2.6-Flash is an open-source foundation model developed by Xiaomi. Built on a Mixture-of-Experts architecture with 309B total parameters and 15B activated per token, it employs a hybrid attention mechanism for...","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"MiMo-V2.6-Pro is the flagship foundation model developed by Xiaomi. Built at a scale of over 1T parameters, it is designed to push the ceiling of capability for the most demanding...","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.003,"tiers":[{"input":0.8,"output":4,"cache_read":0.16,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.8,"output":4,"cache_read":0.16}},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo-V2.6-Pro-UltraSpeed is the fast speed edition of Xiaomi's flagship foundation model, MiMo-V2.6-Pro. Built from the same 1T MiMo-V2.6-Pro checkpoint, it matches the original model in quality while delivering roughly 10x...","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.435,"output":0.87,"cache_read":0.004,"tiers":[{"input":2,"output":6,"cache_read":0.4,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.4}},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"perceptron/perceptron-mk1":{"id":"perceptron/perceptron-mk1","name":"Perceptron: Perceptron Mk1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"perceptron/perceptron-mk1"},"perceptron/perceptron-mk1.5":{"id":"perceptron/perceptron-mk1.5","name":"Perceptron: Perceptron Mk1.5","description":"Perceptron Mk1.5 is Perceptron's embodied reasoning model for physical agents. It accepts text, image, video, and audio input, and answers with text plus optional structured annotations: points, boxes, polygons, tracks,...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":36864,"output":8192},"cost":{"input":0.15,"output":1.5}},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Upstage: Solar Pro 4","description":"Solar Pro 4 is a large language model from Upstage. It is suited for agentic workflows, office productivity, document-intensive work, and coding.","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Upstage: Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":117964},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-mini4":{"id":"upstage/solar-mini4","name":"Upstage: Solar Mini 4","description":"Solar Mini 4 is Upstage's compact, cost-efficient language model, a 35B-parameter mixture-of-experts with 3B active parameters and a 524K context window. It is built for agentic use cases where response...","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"upstage/solar-mini4"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.125,"output":0.75,"reasoning":0.75,"cache_read":0.0125,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.04,"output":0.12},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.625,"output":5,"reasoning":5,"cache_read":0.0625,"cache_write":0.1875},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":1,"output":6,"reasoning":6,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/lyria-3-clip-preview":{"id":"google/lyria-3-clip-preview","name":"Lyria 3 Clip Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-clip-preview"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Gemini 3.6 Flash is a high-efficiency model from Google for coding, agentic workflows, and web and app development. It is designed to produce polished outputs with fewer unnecessary edits and...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.375,"output":1.875,"reasoning":1.875,"cache_read":0.0375,"cache_write":0.020833},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemma-2-27b-it":{"id":"google/gemma-2-27b-it","name":"Gemma 2 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-07-13","last_updated":"2024-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0.65,"output":0.65},"canonical_model_id":"google/gemma-2-27b-it"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.042,"output":0.22,"cache_read":0.021},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Gemini 3.5 Flash Lite is a high-efficiency model from Google with upgraded agentic capabilities. It is suited for subagents that execute focused tasks within complex, multi-agent workflows.","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.15,"output":1.25,"reasoning":1.25,"cache_read":0.015,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/lyria-3-pro-preview":{"id":"google/lyria-3-pro-preview","name":"Lyria 3 Pro Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-pro-preview"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini 4 Argon is Google's flagship reasoning model, built for complex software engineering, deep mathematical and scientific problem solving, and autonomous multi-step agents. Google trained it to keep context across...","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.125,"output":0.75,"reasoning":0.75,"cache_read":0.0125,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Nano Banana 2.1 (Gemini Nano Banana 2.1) is Google's image generation and editing model on the Flash tier, succeeding Nano Banana 2 and Nano Banana Pro. It improves product recontextualization,...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1,"output":6,"reasoning":6,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"Gemini 3.7 Flash is a multimodal model from Google for fast agentic workflows, coding, and complex multi-step reasoning. It is designed for tasks that require responsive performance and reliable multi-step...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.041667},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro is Google’s most advanced image-generation and editing model, built on Gemini 3 Pro. It extends the original Nano Banana with significantly improved multimodal reasoning, real-world grounding, and...","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.15,"output":1.25,"reasoning":1.25,"cache_read":0.015,"cache_write":0.041667},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Gemini 3.1 Flash Image, a.k.a. \"Nano Banana 2,\" is Google’s latest state of the art image generation and editing model, delivering Pro-level visual quality at Flash speed. It combines advanced...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemini-2.5-pro-preview":{"id":"google/gemini-2.5-pro-preview","name":"Gemini 2.5 Pro Preview 06-05","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["pdf","image","text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.625,"output":5,"reasoning":5,"cache_read":0.0625,"cache_write":0.1875},"canonical_model_id":"google/gemini-2.5-pro-preview"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.04,"output":0.05,"cache_read":0.02},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.1,"output":0.34,"cache_read":0.1},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Gemini 3.8 Flash is Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows.","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) is Google's fastest, most cost-efficient Gemini image model, built for high-velocity developer pipelines and rapid-fire visual exploration. It delivers text-to-image generation...","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":4.5,"reasoning":4.5,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.05,"output":0.2,"reasoning":0.2,"cache_read":0.005,"cache_write":0.041667},"canonical_model_id":"google/gemini-2.5-flash-lite"},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Step 3.7 Flash is StepFun's latest high-efficiency multimodal Mixture-of-Experts model. It pairs a 196B-parameter language backbone with a vision encoder for native image and video understanding, activating roughly 11B parameters...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":230400},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"Step 5 Preview is StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters). It performs strongly in software engineering and professional...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":150,"output":600},"canonical_model_id":"openai/o1-pro"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-audio-mini":{"id":"openai/gpt-audio-mini","type":"chat","name":"GPT-Audio Mini","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio","pdf"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":2.4},"canonical_model_id":"openai/gpt-audio-mini"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra","description":"GPT-5.6 Terra Pro is the same underlying model as [GPT-5.6 Terra](https://openrouter.ai/openai/gpt-5.6-terra), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"GPT-5.6 Terra is a balanced model in OpenAI's GPT-5.6 series, positioned between the flagship Sol tier and the cost-efficient Luna tier. It is suited for everyday coding, reasoning, and agentic...","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"GPT-6.1 Sol is an upgrade to GPT-6 Sol from OpenAI, positioned below the flagship GPT-6 Astra in the GPT-6 series. It is suited for agentic coding, computer use, document-heavy professional...","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6.1-sol-pro":{"id":"openai/gpt-6.1-sol-pro","name":"OpenAI: GPT-6.1 Sol Pro","description":"GPT-6.1 Sol Pro is the same underlying model as [GPT-6.1 Sol](https://openrouter.ai/openai/gpt-6.1-sol), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. **Cost note:** pro mode spends far more...","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.2-chat":{"id":"openai/gpt-5.2-chat","name":"OpenAI: GPT-5.2 Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.018,"output":0.09,"cache_read":0.009},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"GPT-6 Luna is the fast, cost-efficient model in OpenAI's GPT-6 series, positioned below GPT-6 Sol. It is suited for high-volume and latency-sensitive workloads such as chat, classification, and lightweight agentic...","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"GPT-5.6 Luna is a fast, cost-efficient model in OpenAI's GPT-5.6 series. It is suited for high-volume, latency-sensitive tasks such as chat, classification, and lightweight agentic workflows, providing capable reasoning for...","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-3.5-turbo-16k":{"id":"openai/gpt-3.5-turbo-16k","name":"OpenAI: GPT-3.5 Turbo 16k","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2023-08-28","last_updated":"2023-08-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":3,"output":4}},"openai/gpt-chat-latest":{"id":"openai/gpt-chat-latest","name":"OpenAI: GPT Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5}},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":4096},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/o4-mini-high":{"id":"openai/o4-mini-high","name":"OpenAI: o4 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275}},"openai/gpt-3.5-turbo-instruct":{"id":"openai/gpt-3.5-turbo-instruct","name":"OpenAI: GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-09-28","last_updated":"2023-09-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1.5,"output":2}},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra-pro":{"id":"openai/gpt-6-astra-pro","name":"OpenAI: GPT-6 Astra Pro ($$$$)","description":"GPT-6 Astra Pro is the same underlying model as [GPT-6 Astra](https://openrouter.ai/openai/gpt-6-astra), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon...","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5-image-mini":{"id":"openai/gpt-5-image-mini","type":"image","name":"GPT-5 Image Mini","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["pdf","image","text"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":2,"cache_read":0.25},"canonical_model_id":"openai/gpt-5-image-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-4o-mini-2024-07-18":{"id":"openai/gpt-4o-mini-2024-07-18","name":"OpenAI: GPT-4o-mini (2024-07-18)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol","description":"GPT-5.6 Sol Pro is the same underlying model as [GPT-5.6 Sol](https://openrouter.ai/openai/gpt-5.6-sol), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-6-sol-pro":{"id":"openai/gpt-6-sol-pro","name":"OpenAI: GPT-6 Sol Pro","description":"GPT-6 Sol Pro is the same underlying model as [GPT-6 Sol](https://openrouter.ai/openai/gpt-6-sol), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-6-luna-pro":{"id":"openai/gpt-6-luna-pro","name":"OpenAI: GPT-6 Luna Pro","description":"GPT-6 Luna Pro is the same underlying model as [GPT-6 Luna](https://openrouter.ai/openai/gpt-6-luna), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-audio":{"id":"openai/gpt-audio","type":"chat","name":"GPT-Audio","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","audio","pdf"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-audio"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.03,"output":0.15,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","pdf","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-5.4-image-2":{"id":"openai/gpt-5.4-image-2","type":"image","name":"GPT-5.4 Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":8,"output":15,"cache_read":2},"canonical_model_id":"openai/gpt-5.4-image-2"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/o3":{"id":"openai/o3","name":"o3","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna","description":"GPT-5.6 Luna Pro is the same underlying model as [GPT-5.6 Luna](https://openrouter.ai/openai/gpt-5.6-luna), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"GPT-5.6 Sol is the flagship model in OpenAI's GPT-5.6 series. It is suited for complex reasoning, coding, and agentic workflows, and is particularly strong at command-line and multi-step coding tasks...","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. It is suited for demanding professional...","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5-image":{"id":"openai/gpt-5-image","type":"image","name":"GPT-5 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-10-01","release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":10,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-5-image"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-3.5-turbo-0613":{"id":"openai/gpt-3.5-turbo-0613","name":"OpenAI: GPT-3.5 Turbo (older v0613)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1,"output":2}},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"OpenAI: o3 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-02-12","last_updated":"2025-02-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"amazon/nova-premier-v1":{"id":"amazon/nova-premier-v1","name":"Amazon: Nova Premier 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-30","last_updated":"2025-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"status":"deprecated","cost":{"input":2.5,"output":12.5,"cache_read":0.625},"canonical_model_id":"amazon/nova-premier"},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Amazon: Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"amazon/nova-lite-v1"},"amazon/nova-2-lite-v1":{"id":"amazon/nova-2-lite-v1","name":"Amazon: Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"amazon/nova-2-lite-v1"},"amazon/nova-micro-v1":{"id":"amazon/nova-micro-v1","name":"Amazon: Nova Micro 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":5120},"cost":{"input":0.035,"output":0.14},"canonical_model_id":"amazon/nova-micro-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Amazon: Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.8,"output":3.2},"canonical_model_id":"amazon/nova-pro-v1"},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Perplexity: Sonar Pro","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"perplexity/sonar-reasoning-pro":{"id":"perplexity/sonar-reasoning-pro","name":"Perplexity: Sonar Reasoning Pro","description":"Web-grounded reasoning model for multi-step research and cited answers","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8}},"perplexity/sonar":{"id":"perplexity/sonar","name":"Perplexity: Sonar","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-27","last_updated":"2025-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127072,"output":114364},"cost":{"input":1,"output":1}},"perplexity/sonar-deep-research":{"id":"perplexity/sonar-deep-research","name":"Perplexity: Sonar Deep Research","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar-deep-research","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8,"reasoning":3}},"perplexity/sonar-pro-search":{"id":"perplexity/sonar-pro-search","name":"Perplexity: Sonar Pro Search","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-30","last_updated":"2025-10-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"bytedance/ui-tars-1.5-7b":{"id":"bytedance/ui-tars-1.5-7b","name":"ByteDance: UI-TARS 7B ","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":2048},"cost":{"input":0.1,"output":0.2,"cache_read":0.1}},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"MoonshotAI: Kimi K2.7 Code is a coding-focused model in Moonshot AI's Kimi K2 family, built to complete end-to-end programming tasks reliably over long contexts. It uses a native multimodal mixture-of-experts...","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"MoonshotAI: Kimi K3.1 Pro","description":"Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. It is built for long-horizon coding, 3D...","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.65,"output":3.41,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 Instruct 0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Meta: Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 Contributor is the cost-efficient contributor tier of Meta’s multimodal reasoning model for experimentation, learning, and early-stage agentic, multi-agent, and coding workflows. It is designed to track information...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark 1.1 is a multimodal reasoning model from Meta, built for agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with a 1M-token context...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a reasoning model from Meta, designed for complex agentic tasks. It accepts text, images, video, audio, and PDF documents, returns text, and offers a 1M-token context...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is a coding-optimized open-weight model. Trained specifically for agentic coding workflows, it delivers higher first-attempt accuracy, reliable tool calling, and deep codebase understanding across the repos developers...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer 30B is a dense, open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark and optimized for autonomous agents on consumer hardware. It is suited for long-horizon...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It is designed to keep track of information across extended tasks, work through...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-wonder":{"id":"meta/muse-wonder","name":"Meta: Muse Wonder","description":"Muse Wonder is a frontier reasoning model from Meta, designed for complex personal and professional agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":4,"output":20,"cache_read":0.2}},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Meta: Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 contributor tier is a reasoning model from Meta designed for developers who want to start building at an even lower cost. It’s meaningfully cheaper than Muse Spark...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-08-21","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":3686},"cost":{"input":0.08,"output":0.11},"canonical_model_id":"gryphe/mythomax-l2-13b"},"~google/gemini-flash-latest":{"id":"~google/gemini-flash-latest","name":"Google: Gemini Flash Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667}},"~google/gemini-pro-latest":{"id":"~google/gemini-pro-latest","name":"Google: Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["audio","pdf","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875}},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 25.08","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":204800},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral Nemo","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-large-2407":{"id":"mistralai/mistral-large-2407","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-2407"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Mistral Large 3 2512 is Mistral’s most capable model to date, featuring a sparse mixture-of-experts architecture with 41B active parameters (675B total), and released under the Apache 2.0 license.","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":26214},"cost":{"input":0.2,"output":0.6,"cache_read":0.02},"canonical_model_id":"mistral/mistral-saba"},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2","description":"Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window. Devstral 2 supports exploring...","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3b-2512"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Mistral Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":102400},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-medium-3-5":{"id":"mistralai/mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistralai/mistral-large-4-0":{"id":"mistralai/mistral-large-4-0","name":"Mistral: Mistral Large 4","description":"Mistral Large 4 is a frontier multimodal (text and image input) model from Mistral AI built for reasoning, coding, and agentic workloads. It offers a 512K-token context window with up...","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Mistral Small 3.2 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Mistral Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":102400},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-8b-2512"},"mistralai/voxtral-small-24b-2507":{"id":"mistralai/voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":26214},"cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It is suited for coding, professional workflows, research, multimodal interaction, and long-running agent tasks, with flexible thinking that can be...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.425,"output":2.55,"cache_read":0.085,"cache_write":0.53125},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen2.5-vl-72b-instruct":{"id":"qwen/qwen2.5-vl-72b-instruct","name":"Qwen: Qwen2.5 VL 72B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":115200},"cost":{"input":0.8,"output":1,"cache_read":0.4},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"qwen/qwen3-235b-a22b-2507":{"id":"qwen/qwen3-235b-a22b-2507","name":"Qwen: Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.5-plus-20260420":{"id":"qwen/qwen3.5-plus-20260420","name":"Qwen: Qwen3.5 Plus 2026-04-20","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.8,"cache_write":0.375}},"qwen/qwen3.8-max-prime":{"id":"qwen/qwen3.8-max-prime","name":"Qwen 3.8 Max Prime","description":"Qwen3.8 Max Prime is a higher-throughput variant of Qwen3.8 Max from Alibaba's Qwen team, served as a separate SKU at a higher price point. It accepts text, image, and video...","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"qwen/qwen3-vl-30b-a3b-thinking":{"id":"qwen/qwen3-vl-30b-a3b-thinking","name":"Qwen: Qwen3 VL 30B A3B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.29,"output":1}},"qwen/qwen3.5-plus-02-15":{"id":"qwen/qwen3.5-plus-02-15","name":"Qwen: Qwen3.5 Plus 2026-02-15","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.26,"output":1.56}},"qwen/qwen3.5-flash-02-23":{"id":"qwen/qwen3.5-flash-02-23","name":"Qwen: Qwen3.5-Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-25","last_updated":"2026-02-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.065,"output":0.26}},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.21,"output":1.9,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"cache_write":0.40625},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-coder-30b-a3b-instruct":{"id":"qwen/qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen3 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Qwen3.7 Flash is a vision-language reasoning model from Alibaba. It is suited for multimodal agents, visual coding, search, and computer interaction, with strengths in object recognition, spatial understanding, and real-world...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.195,"output":0.975,"cache_read":0.039,"cache_write":0.24375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.1625,"output":1.3},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-vl-8b-instruct":{"id":"qwen/qwen3-vl-8b-instruct","name":"Qwen: Qwen3 VL 8B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.25,"output":0.75,"cache_read":0.12}},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"status":"deprecated","cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen: Qwen3 235B A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.45,"output":3.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen3.8 Omni Flash is an omni-modal reasoning model from Alibaba, the first Qwen model built around agentic capabilities with native audio-video understanding. It is suited for audio-video analysis and summarization,...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26,"output":2.08},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen: Qwen3 30B A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.09,"output":0.3},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.39,"output":2.34},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen-2.5-72b-instruct":{"id":"qwen/qwen-2.5-72b-instruct","name":"Qwen2.5 72B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"cost":{"input":0.36,"output":0.4}},"qwen/qwen-2.5-7b-instruct":{"id":"qwen/qwen-2.5-7b-instruct","name":"Qwen: Qwen2.5 7B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-10-16","last_updated":"2024-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":29491},"cost":{"input":0.1,"output":0.2}},"qwen/qwen3-14b":{"id":"qwen/qwen3-14b","name":"Qwen: Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.24}},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"Qwen3.8 Max 0902 is an updated snapshot of Qwen3.8 Max from Alibaba's Qwen team. It is a 2.4-trillion-parameter mixture-of-experts model that accepts text, image, and video input and returns text,...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":909000},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.45,"output":2.7},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen3.8 Flash is a multimodal reasoning model from Alibaba. It is suited for coding assistance, agentic workflows, visual understanding, document and codebase analysis, desktop interaction, chart analysis, and long-video analysis.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.26,"output":0.78,"cache_read":0.052,"cache_write":0.325},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Qwen3.7-Plus is a cost-effective model in Alibaba's Qwen3.7 series. It supports text and image input with text output, building on the series' text capabilities with a comprehensive upgrade to its...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.32,"output":1.28,"cache_read":0.064,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen-2.5-coder-32b-instruct":{"id":"qwen/qwen-2.5-coder-32b-instruct","name":"Qwen2.5 Coder 32B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.195,"output":1.56},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.12,"output":0.8,"cache_read":0.07},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.1,"output":0.95,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1875,"output":1.125,"cache_write":0.234375},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3-coder":{"id":"qwen/qwen3-coder","name":"Qwen: Qwen3 Coder 480B A35B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":1,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"~deepseek/deepseek-v4-flash-latest":{"id":"~deepseek/deepseek-v4-flash-latest","name":"DeepSeek: DeepSeek V4 Flash Latest","description":"This model always redirects to the latest model in the DeepSeek V4 Flash family.","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-01","last_updated":"2026-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.0422,"output":0.1729,"cache_read":0.027}},"~deepseek/deepseek-flash-latest":{"id":"~deepseek/deepseek-flash-latest","name":"DeepSeek: DeepSeek Flash Latest","description":"This model always redirects to the latest model in the DeepSeek Flash family.","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.02,"output":1,"cache_read":0.015}},"~deepseek/deepseek-pro-latest":{"id":"~deepseek/deepseek-pro-latest","name":"DeepSeek: DeepSeek Pro Latest","description":"This model always redirects to the latest model in the DeepSeek Pro family.","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":393216},"cost":{"input":0.1188,"output":8,"cache_read":0.12}},"elorian/glimpse":{"id":"elorian/glimpse","name":"Elorian: Glimpse","description":"Glimpse is a multimodal visual reasoning model from Elorian, built for visual understanding and complex visual reasoning in the physical world. It reasons directly over images and video rather than...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":256000},"cost":{"input":0.6,"output":3,"cache_read":0.06}},"openrouter/auto":{"id":"openrouter/auto","name":"Auto Router","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-15","last_updated":"2026-03-15","modalities":{"input":["audio","image","pdf","text","video"],"output":["image","text"]},"open_weights":false,"limit":{"context":2000000,"output":32768},"cost":{"input":0,"output":0}},"openrouter/bodybuilder":{"id":"openrouter/bodybuilder","name":"Body Builder (beta)","description":"Preview model for early access evaluation, prototyping, and compatibility testing","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-03-15","last_updated":"2026-03-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"status":"beta","cost":{"input":0,"output":0}},"openrouter/pareto-code":{"id":"openrouter/pareto-code","name":"Pareto Code Router","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-05-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":65536},"cost":{"input":0,"output":0}},"openrouter/free":{"id":"openrouter/free","name":"OpenRouter Free Models Router","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-01","last_updated":"2026-02-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32768},"cost":{"input":0,"output":0}},"liquid/lfm-2.5-2.6b:free":{"id":"liquid/lfm-2.5-2.6b:free","name":"LiquidAI: LFM2.5-2.6B (free)","description":"LFM2.5-2.6B is a compact reasoning model from Liquid AI. It is suited for agent workflows, data extraction, RAG, and long-context processing. Liquid advises against using it for agentic coding or...","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"cost":{"input":0,"output":0}},"liquid/lfm-2.5-24b-a2b":{"id":"liquid/lfm-2.5-24b-a2b","name":"LiquidAI: LFM2.5-24B-A2B","description":"LFM2.5-24B-A2B is a mixture-of-experts reasoning model from Liquid AI with 24B total and 2.45B active parameters. It always reasons before answering and is built for agents that call tools. Liquid...","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"thedrummer/skyfall-36b-v2":{"id":"thedrummer/skyfall-36b-v2","name":"TheDrummer: Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.55,"output":0.8,"cache_read":0.25},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"thedrummer/unslopnemo-12b":{"id":"thedrummer/unslopnemo-12b","name":"TheDrummer: UnslopNemo 12B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-11-08","last_updated":"2024-11-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":819200},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"thedrummer/unslopnemo-12b"},"thedrummer/cydonia-24b-v4.1":{"id":"thedrummer/cydonia-24b-v4.1","name":"TheDrummer: Cydonia 24B V4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.3,"output":0.5,"cache_read":0.15},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"writer/palmyra-x5":{"id":"writer/palmyra-x5","name":"Writer: Palmyra X5","description":"General-purpose chat model for instruction following, writing, and analysis","family":"palmyra","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"output":8192},"cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"stealth/claude-opus-4.6":{"id":"stealth/claude-opus-4.6","name":"Stealth: Claude Opus 4.6 (20% off)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"reasoning":0,"cache_read":0.4,"cache_write":5}},"stealth/qwen3.6-plus":{"id":"stealth/qwen3.6-plus","name":"Stealth: Qwen3.6 Plus (50% off)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. This third-party-served variant of Qwen3.6 Plus is offered at 50% lower cost than standard Qwen3.6 Plus pricing and is not served by Alibaba or Kilo Code. Note: a surcharge applies to long-context workloads exceeding 256K input tokens.","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":0,"cache_read":0.025,"cache_write":0.3125}},"stealth/claude-sonnet-4.6":{"id":"stealth/claude-sonnet-4.6","name":"Stealth: Claude Sonnet 4.6 (20% off)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.4,"output":12,"reasoning":0,"cache_read":0.24,"cache_write":3}},"stealth/claude-opus-4.7":{"id":"stealth/claude-opus-4.7","name":"Stealth: Claude Opus 4.7 (20% off)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"reasoning":0,"cache_read":0.4,"cache_write":5}},"stealth/glyph-cluster":{"id":"stealth/glyph-cluster","name":"Stealth: Glyph Cluster (free)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. Glyph Cluster is a reasoning model for coding and long-context analysis, served through Vercel AI Gateway during its stealth period at no cost. It supports function calling and streaming, takes text input only, and does not support structured outputs. Zero Data Retention is not available for this model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0,"output":0,"reasoning":0,"cache_read":0}},"stealth/claude-opus-4.8":{"id":"stealth/claude-opus-4.8","name":"Stealth: Claude Opus 4.8 (20% off)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. This third-party-served variant of Claude Opus 4.8 is offered at 20% lower cost than standard Claude Opus 4.8 pricing and is not served by Anthropic or Kilo Code.","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"reasoning":0,"cache_read":0.4,"cache_write":5}},"~moonshotai/kimi-latest":{"id":"~moonshotai/kimi-latest","name":"MoonshotAI: Kimi Latest","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.238,"output":24,"cache_read":0.228}},"nousresearch/hermes-3-llama-3.1-405b":{"id":"nousresearch/hermes-3-llama-3.1-405b","name":"Nous: Hermes 3 405B Instruct (retires Nov 2)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":1,"output":1},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"nousresearch/hermes-3-llama-3.1-70b":{"id":"nousresearch/hermes-3-llama-3.1-70b","name":"Nous: Hermes 3 70B Instruct (retires Nov 2)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":0.7},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"nousresearch/hermes-4-405b":{"id":"nousresearch/hermes-4-405b","name":"Nous: Hermes 4 405B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":1,"output":3},"canonical_model_id":"nousresearch/hermes-4-405b"},"mancer/weaver":{"id":"mancer/weaver","name":"Mancer: Weaver (alpha)","description":"An attempt to recreate Claude-style verbosity, but don't expect the same level of coherence or memory. Meant for use in roleplay/narrative situations.","family":"alpha","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-08-02","last_updated":"2023-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":6000},"cost":{"input":0.4,"output":0.75},"canonical_model_id":"mancer/weaver"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort:...","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.0825,"output":0.33,"cache_read":0.020625},"canonical_model_id":"tencent/hy3"},"tencent/hy3-preview":{"id":"tencent/hy3-preview","name":"Hy3 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.18,"output":0.6,"cache_read":0.06},"canonical_model_id":"tencent/hy3-preview"},"tencent/hy-mt2-7b":{"id":"tencent/hy-mt2-7b","name":"Hy-MT2-7B","description":"Hy-MT2-7B is a 7B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided translation.","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-7b"},"tencent/hy-mt2-1.8b":{"id":"tencent/hy-mt2-1.8b","name":"Hy-MT2-1.8B","description":"Hy-MT2-1.8B is a compact 1.8B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided...","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.044,"output":0.177},"canonical_model_id":"tencent/hy-mt2-1.8b"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"Tencent: Hy4 preview is a mixture-of-experts model from Tencent, with 49B active parameters out of 770B total. It is designed for coding agents, complex tool-use workflows, and productivity tasks that...","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.7506,"output":2.2509,"cache_read":0.0378},"canonical_model_id":"tencent/hy4-preview"},"tencent/hunyuan-a13b-instruct":{"id":"tencent/hunyuan-a13b-instruct","name":"Tencent: Hunyuan A13B Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":117964},"cost":{"input":0.14,"output":0.57}},"tencent/hy-mt2-30b-a3b":{"id":"tencent/hy-mt2-30b-a3b","name":"Hy-MT2-30B-A3B","description":"Hy-MT2-30B-A3B is Tencent's flagship translation model in the Hy-MT2 family. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and...","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-30b-a3b"},"nex-agi/nex-n2.5-mini":{"id":"nex-agi/nex-n2.5-mini","name":"Nex AGI: Nex-N2.5-Mini","description":"Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. Its core strength is agentic coding within a visual feedback loop: it can explore codebases, implement multi-file...","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.025,"output":0.1,"cache_read":0.0025}},"nex-agi/nex-n2.5-pro":{"id":"nex-agi/nex-n2.5-pro","name":"Nex AGI: Nex-N2.5-Pro","description":"Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. Its core strength is agentic coding within a visual feedback loop: it can explore codebases, implement multi-file...","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.075,"output":0.25,"cache_read":0.015}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Inference.net: Schematron V2 Turbo","description":"Schematron V2 Turbo is a 3B-parameter HTML-to-JSON extraction model from Inference.net. It prioritizes throughput for high-volume extraction workloads. Extraction instructions must be supplied through a JSON schema in response_format rather...","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Inference.net: Schematron V2 Small","description":"Schematron V2 Small is a 3B-parameter HTML-to-JSON extraction model from Inference.net. It prioritizes extraction quality for complex schemas and long pages. Extraction instructions must be supplied through a JSON schema...","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.05},"canonical_model_id":"inference-net/schematron-v2-small"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Claude Haiku 5.5 is Anthropic's small, fast model for high-volume, cost-sensitive work such as summarization, subagents, and browser use. It succeeds Claude Haiku 4.5 with stronger coding, computer use, and...","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It is particularly strong at multi-step changes in large codebases, code...","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude Fable 5 is a Mythos-class model from Anthropic, built for autonomous knowledge work and coding. It supports text, image, and file inputs with text output, with reasoning support and...","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Claude Sonnet 5.5 is Anthropic's Sonnet-class model for well-scoped everyday work, succeeding Claude Sonnet 5 as a direct upgrade. It is especially strong at building features, fixing bugs, and producing...","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Claude Opus 5 is Anthropic’s flagship model for demanding reasoning, coding, and long-horizon agentic work. It is particularly strong at end-to-end software tasks, code review and bug finding, visual analysis...","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Sonnet 5 is Anthropic's most capable Sonnet-class model, with frontier performance across coding, agents, and professional work. It supports adaptive thinking with selectable reasoning effort levels (low, medium, high, max,...","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Claude Opus 4.8 is Anthropic's most capable generally available model in the Opus family. It supports text, image, and file inputs with text output, with reasoning support and a 1M-token...","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude Fable 5.1 improves on Claude Fable 5 across the board, with the biggest gains in agentic coding, long-running agentic workflows, and knowledge work: long code refactors, front-end and visual...","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3-prime":{"id":"z-ai/glm-5.3-prime","name":"Z.ai: GLM 5.3 Prime","description":"GLM-5.3-Prime is the high-speed variant of Z.ai's GLM-5.3, inheriting its full capabilities while delivering 1.5–2× the output throughput through inference acceleration. It supports text input and output with a 1M-token...","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.56}},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves...","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"Z.ai: GLM 5.3 FlashX","description":"GLM-5.3-FlashX is the high-speed variant of Z.ai's GLM-5.3-Flash, a native multimodal model delivering inference speeds of up to 200 tokens/s. Built on the same hybrid sparse and linear attention architecture...","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.09}},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.54,"output":1.98,"cache_read":0.099},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM-4.5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"GLM 5.2 is a large-scale reasoning model from Z.ai. It supports text input and output with a 1M-token context window, and is suited for long-horizon agent workflows, project-level software engineering,...","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"GLM-5.3-Flash is a native multimodal model from Z.ai. It is suited for efficient coding and long-horizon agent tasks. Its hybrid sparse and linear attention architecture maintains accurate long-context behavior while...","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":943717},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM-4.5-Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":2.5,"output":5},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"meta-llama/llama-4-scout":{"id":"meta-llama/llama-4-scout","name":"Meta: Llama 4 Scout","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"output":16384},"cost":{"input":0.08,"output":0.24},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama-3.1-8B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.02,"output":0.04},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-guard-4-12b":{"id":"meta-llama/llama-guard-4-12b","name":"Meta: Llama Guard 4 12B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0.144,"output":0.144},"canonical_model_id":"meta/llama-guard-4-12b"},"meta-llama/llama-3.2-3b-instruct":{"id":"meta-llama/llama-3.2-3b-instruct","name":"Meta: Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":72000},"cost":{"input":0.05,"output":0.33},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"meta-llama/llama-3.1-70b-instruct":{"id":"meta-llama/llama-3.1-70b-instruct","name":"Llama-3.1-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"meta-llama/llama-4-maverick":{"id":"meta-llama/llama-4-maverick","name":"Meta: Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1875,"output":0.6525,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/llama-3.2-1b-instruct":{"id":"meta-llama/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":54000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"nvidia/nemotron-3.5-super-vl:free":{"id":"nvidia/nemotron-3.5-super-vl:free","name":"NVIDIA: Nemotron 3.5 Super VL (free)","description":"NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. It takes text, images, documents, GUI screenshots, and video... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3.5-content-safety:free":{"id":"nvidia/nemotron-3.5-content-safety:free","name":"NVIDIA: Nemotron 3.5 Content Safety (free)","description":"NVIDIA Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model from NVIDIA, fine-tuned from Google Gemma-3-4B. It moderates both inputs to and responses from LLMs and VLMs, accepting... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"NVIDIA: Nemotron 3.5 Super VL","description":"NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. It takes text, images, documents, GUI screenshots, and video...","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.1,"output":0.4,"cache_read":0.08},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3-super-120b-a12b:free":{"id":"nvidia/nemotron-3-super-120b-a12b:free","name":"NVIDIA: Nemotron 3 Super (free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nemotron 3.5 Lightning 30B A3B","description":"NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. It is suited for high-throughput agentic workloads and specialized tasks that...","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.025,"output":0.12,"cache_read":0.0125},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":182520},"cost":{"input":0.4,"output":1.76,"cache_read":0.08},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free":{"id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free","name":"NVIDIA: Nemotron 3 Nano Omni (free)","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"NVIDIA Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model from NVIDIA, fine-tuned from Google Gemma-3-4B. It moderates both inputs to and responses from LLMs and VLMs, accepting...","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3.5-lightning:free":{"id":"nvidia/nemotron-3.5-lightning:free","name":"NVIDIA: Nemotron 3.5 Lightning (free)","description":"NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. It is suited for high-throughput agentic workloads and specialized tasks that... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.08,"output":0.45},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b:free":{"id":"nvidia/nemotron-3-ultra-550b-a55b:free","name":"NVIDIA: Nemotron 3 Ultra (free)","description":"NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"~openai/gpt-mini-latest":{"id":"~openai/gpt-mini-latest","name":"OpenAI: GPT Mini Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075}},"~openai/gpt-astra-latest":{"id":"~openai/gpt-astra-latest","name":"OpenAI: GPT Astra Latest ($$$$)","description":"This model always redirects to the latest model in the OpenAI GPT Astra family.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"~openai/gpt-luna-latest":{"id":"~openai/gpt-luna-latest","name":"OpenAI: GPT Luna Latest","description":"This model always redirects to the latest model in the OpenAI GPT Luna family.","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"~openai/gpt-sol-latest":{"id":"~openai/gpt-sol-latest","name":"OpenAI: GPT Sol Latest","description":"This model always redirects to the latest model in the OpenAI GPT Sol family.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"~openai/gpt-terra-latest":{"id":"~openai/gpt-terra-latest","name":"OpenAI: GPT Terra Latest","description":"This model always redirects to the latest model in the OpenAI GPT Terra family.","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5}},"meituan/longcat-2.0":{"id":"meituan/longcat-2.0","name":"Meituan: LongCat 2.0","description":"LongCat 2.0 is a sparse mixture-of-experts language model from Meituan, with 48B active parameters out of 1.6T total. It is suited for coding, repository-level changes, long-horizon problem solving, and agentic...","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048756,"output":262144},"cost":{"input":0.75,"output":3,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"xingchen-agi/xing4.0-29b-a4b:free":{"id":"xingchen-agi/xing4.0-29b-a4b:free","name":"XingChen: Xing4.0 29B A4B (free)","description":"Xing4.0-29B-A4B is an open-weight mixture-of-experts language model from China Telecom's AI team (TeleAI) and the successor to the TeleChat series. It has 29B total parameters with 4B active per token...","family":"agi","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"unbiased/pareto-26.10":{"id":"unbiased/pareto-26.10","name":"Pareto 26.10","description":"Pareto is a multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of general-purpose tasks.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"unbiased/pareto-26.9":{"id":"unbiased/pareto-26.9","name":"Pareto 26.9","description":"Pareto is a multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of general-purpose tasks.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25}},"prism-ml/ternary-bonsai-2-27b":{"id":"prism-ml/ternary-bonsai-2-27b","name":"PrismML: Ternary Bonsai 2 27B","description":"Bonsai 2 27B is a 27B-parameter reasoning model from PrismML derived from Qwen3.8-27B. It supports coding, mathematics, tool calling, and image understanding with a 262K-token context window. Ternary compression shrinks...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.5,"cache_read":0.0375}},"recursive/flint-v1.0:free":{"id":"recursive/flint-v1.0:free","name":"Recursive: Flint (free)","description":"Flint is an open-weight mixture-of-experts model from Recursive, with about 2.5B active parameters out of 40B total, built by Recursive's autoresearch system Eureka Machine for coding agents. It is suited...","family":"flint","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Fireworks: Ember-1","description":"Ember-1 is a specialized reasoning model from Fireworks Research, built on [Kimi K3](https://openrouter.ai/moonshotai/kimi-k3). It is designed to make every token go further: it produces shorter reasoning traces, using roughly 40%...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"dots-studio/dots-3-note-preview:free":{"id":"dots-studio/dots-3-note-preview:free","name":"Dots Studio: Dots3-Note Preview (free)","description":"Dots3-Note Preview is an open-weight mixture-of-experts model from Dots Studio, with 16B active parameters out of 280B total. It is the lightest model in the Dots 3 family and is...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":460800},"cost":{"input":0,"output":0}},"kilo-auto/balanced":{"id":"kilo-auto/balanced","name":"Auto Balanced","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"reasoning":0,"cache_read":0.0325,"cache_write":0.40625}},"kilo-auto/small":{"id":"kilo-auto/small","name":"Auto Small","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.4,"reasoning":0,"cache_read":0.005}},"kilo-auto/frontier":{"id":"kilo-auto/frontier","name":"Auto Frontier","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"reasoning":0,"cache_read":0.5,"cache_write":6.25}},"kilo-auto/efficient":{"id":"kilo-auto/efficient","name":"Auto Efficient","description":"Routes each request to the cheapest model that gets the job done, based on continuously benchmarked accuracy and cost.","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"reasoning":0,"cache_read":0.0325,"cache_write":0.40625}},"kilo-auto/free":{"id":"kilo-auto/free","name":"Auto Free","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"cost":{"input":0,"output":0,"reasoning":0,"cache_read":0,"cache_write":0}},"~unbiased/pareto-latest":{"id":"~unbiased/pareto-latest","name":"Pareto Latest","description":"This model always redirects to the latest model in the Pareto family.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"rekaai/reka-flash-3":{"id":"rekaai/reka-flash-3","name":"Reka Flash 3","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"reka","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":58982},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"rekaai/reka-flash-3"},"rekaai/reka-edge":{"id":"rekaai/reka-edge","name":"Reka Edge","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":14745},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"rekaai/reka-edge"}}},"xpersona":{"id":"xpersona","env":["XPERSONA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://www.xpersona.co/v1","name":"Xpersona","doc":"https://www.xpersona.co/docs","models":{"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":372000,"output":128000},"cost":{"input":1.5,"output":2,"reasoning":2,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.6-terra"},"xpersona-frieren-coder":{"id":"xpersona-frieren-coder","name":"Xpersona Frieren 1","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-30","release_date":"2026-05-01","last_updated":"2026-05-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":384000},"cost":{"input":1.5,"output":6,"reasoning":6,"cache_read":0.15}},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":3.7,"reasoning":3.7,"cache_read":0.06},"canonical_model_id":"anthropic/claude-haiku-4-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.75,"output":6,"reasoning":6,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":18.5,"reasoning":18.5,"cache_read":0.3},"canonical_model_id":"anthropic/claude-fable-5"},"gpt-5.6":{"id":"gpt-5.6","name":"GPT-5.6","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":372000,"output":128000},"cost":{"input":1.5,"output":12,"reasoning":12,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.6-sol"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.5,"output":9.25,"reasoning":9.25,"cache_read":0.15},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":1.5,"output":12,"reasoning":12,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.9,"output":5.55,"reasoning":5.55,"cache_read":0.09},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":372000,"output":128000},"cost":{"input":1.5,"output":12,"reasoning":12,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.55,"output":12.2,"reasoning":12.2,"cache_read":0.155},"canonical_model_id":"google/gemini-3.5-flash"},"xpersona-gpt-5.5":{"id":"xpersona-gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-30","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":3,"output":18,"reasoning":18,"cache_read":0.3},"canonical_model_id":"openai/gpt-5.5"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":0.375,"output":4,"reasoning":4,"cache_read":0.0375},"canonical_model_id":"openai/gpt-5.4-mini"}}},"sambanova":{"id":"sambanova","env":["SAMBANOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.sambanova.ai/v1","name":"SambaNova","doc":"https://docs.sambanova.ai/cloud/docs/get-started/supported-models","models":{"gemma-4-31B-it":{"id":"gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.38,"output":1.15},"canonical_model_id":"google/gemma-4-31b-it"},"DeepSeek-V3.1":{"id":"DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":7168},"cost":{"input":3,"output":4.5},"canonical_model_id":"deepseek/deepseek-v3.1"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.6,"output":2.4,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"DeepSeek-V3.2":{"id":"DeepSeek-V3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":7168},"cost":{"input":3,"output":4.5},"canonical_model_id":"deepseek/deepseek-v3.2"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.22,"output":0.59},"canonical_model_id":"openai/gpt-oss-120b"},"Meta-Llama-3.3-70B-Instruct":{"id":"Meta-Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":3072},"cost":{"input":0.6,"output":1.2},"canonical_model_id":"meta/llama-3.3-70b-instruct"}}},"inco":{"id":"inco","env":["INCO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inco.ai/v1","name":"Inco","doc":"https://platform.inco.ai/docs","models":{"deepseek-v4.1-flash:fast":{"id":"deepseek-v4.1-flash:fast","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.6,"output":2.4,"cache_read":0.012},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3:fast":{"id":"kimi-k3:fast","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":6,"output":30,"cache_read":0.6},"canonical_model_id":"moonshotai/kimi-k3"},"glm-5.3-flash:fast":{"id":"glm-5.3-flash:fast","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1,"cache_read":0.06},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5.3:fast":{"id":"glm-5.3:fast","name":"GLM-5.3 Fast","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.52},"canonical_model_id":"zhipuai/glm-5.3"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax-m3:fast":{"id":"minimax-m3:fast","name":"MiniMax M3 Fast","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"}}},"cloudflare-ai-gateway":{"id":"cloudflare-ai-gateway","env":["CLOUDFLARE_API_TOKEN","CLOUDFLARE_ACCOUNT_ID","CLOUDFLARE_GATEWAY_ID"],"npm":"ai-gateway-provider","name":"Cloudflare AI Gateway","doc":"https://developers.cloudflare.com/ai-gateway/","models":{"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"alibaba/qwen3.7-max":{"id":"alibaba/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":3.75,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.7-max"},"alibaba/qwen3-max":{"id":"alibaba/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"alibaba/qwen3.5-397b-a17b":{"id":"alibaba/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"alibaba/qwen3.7-plus":{"id":"alibaba/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.32,"output":1.28,"cache_read":0.064},"canonical_model_id":"alibaba/qwen3.7-plus"},"minimax/m3.1-flash":{"id":"minimax/m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015,"cache_write":0.15},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"reflection-ai/beam-1.0-501b-a23b":{"id":"reflection-ai/beam-1.0-501b-a23b","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"cache_write":1.5},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-flash":{"id":"google/gemini-3-flash","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro":{"id":"google/gemini-3.1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.25,"output":5,"cache_read":0.625},"canonical_model_id":"openai/gpt-4o"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":10,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.16},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.16},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1040000,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.16},"canonical_model_id":"meta/muse-spark-1.3"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"unbiased/pareto":{"id":"unbiased/pareto","name":"Pareto","description":"Blended multimodal model that runs multiple language models in parallel and returns a synthesized answer","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25},"canonical_model_id":"unbiased/pareto"},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember-1","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"}}},"neosmith":{"id":"neosmith","env":["NEOSMITH_API_KEY"],"npm":"@ai-sdk/openai","api":"https://router.neosmith.ai/v1","name":"NeoSmith","doc":"https://neosmith.ai/docs","models":{"neosmith.intelligent-maestro":{"id":"neosmith.intelligent-maestro","name":"NeoSmith Maestro","description":"Highest-accuracy coding tier. Hard, self-contained problems run NeoSmith's premium multi-model solver; everything else gets the strongest intelligence tier.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.4,"output":12,"cache_read":0.35,"cache_write":0}},"neosmith.intelligent-pro":{"id":"neosmith.intelligent-pro","name":"NeoSmith Pro","description":"Default production tier. Intelligent NeoSmith routing with a Claude Opus ceiling on escalation.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-26","last_updated":"2026-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.81,"output":8.39,"cache_read":0.3,"cache_write":0}},"neosmith.neolite":{"id":"neosmith.neolite","name":"NeoSmith NeoLite","description":"Sealed single-model budget tier. 512K context, text and images, tool use, and no escalation of any kind.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-20","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":64000},"cost":{"input":0.6,"output":2.4,"cache_read":0.08,"cache_write":0}},"neosmith.intelligent-basic":{"id":"neosmith.intelligent-basic","name":"NeoSmith Basic","description":"Cost-capped tier. Intelligent routing with a Claude Sonnet ceiling — Opus is never invoked.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-26","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.17,"output":4.37,"cache_read":0.22,"cache_write":0}}}},"zai-coding-plan":{"id":"zai-coding-plan","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.z.ai/api/coding/paas/v4","name":"Z.AI Coding Plan","doc":"https://docs.z.ai/devpack/overview","models":{"glm-5-turbo":{"id":"glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"glm-5.2-highspeed":{"id":"glm-5.2-highspeed","name":"GLM-5.2 Highspeed","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5.3-highspeed":{"id":"glm-5.3-highspeed","name":"GLM-5.3 Highspeed","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"}}},"wallaby":{"id":"wallaby","env":["WALLABY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.wallabytoken.com/v1","name":"Wallaby","doc":"https://wallabytoken.com/docs","models":{"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.7,"output":13.5,"cache_read":0.27},"canonical_model_id":"moonshotai/kimi-k3"}}},"mixlayer":{"id":"mixlayer","env":["MIXLAYER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://models.mixlayer.ai/v1","name":"Mixlayer","doc":"https://docs.mixlayer.com","models":{"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":1.3},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"}}},"frogbot":{"id":"frogbot","env":["FROGBOT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://app.frogbot.ai/api/v1","name":"FrogBot","doc":"https://docs.frogbot.ai","models":{"qwen-3-6-plus":{"id":"qwen-3-6-plus","name":"Qwen 3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.6-plus"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek v4 Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.31},"canonical_model_id":"google/gemini-2.5-pro"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.2,"output":1.5,"cache_read":0.02},"canonical_model_id":"xai/grok-code-fast-1"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-20b"},"gpt-5-4-mini":{"id":"gpt-5-4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"minimax-m2-5":{"id":"minimax-m2-5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":8192},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"zai-glm-5-1":{"id":"zai-glm-5-1","name":"Z.AI GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":8192},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"gpt-5-5":{"id":"gpt-5-5","name":"GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":8192},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"gemini-3-1-pro-preview":{"id":"gemini-3-1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"grok-4-3":{"id":"grok-4-3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.075},"canonical_model_id":"google/gemini-2.5-flash"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"gpt-5-4-nano":{"id":"gpt-5-4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-5-3-codex":{"id":"gpt-5-3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"}}},"daoxe":{"id":"daoxe","env":["DAOXE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://daoxe.com/v1","name":"DaoXE","doc":"https://daoxe.com/pricing","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":5},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":4,"output":16,"cache_read":0.8},"canonical_model_id":"moonshotai/kimi-k2.5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"}}},"azure-cognitive-services":{"id":"azure-cognitive-services","env":["AZURE_COGNITIVE_SERVICES_RESOURCE_NAME","AZURE_COGNITIVE_SERVICES_API_KEY"],"npm":"@ai-sdk/azure","name":"Azure Cognitive Services","doc":"https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/models","models":{"deepseek-v3.2-speciale":{"id":"deepseek-v3.2-speciale","name":"DeepSeek-V3.2-Speciale","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek-V4.1-Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.71,"output":0.71}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-3.5-turbo-1106":{"id":"gpt-3.5-turbo-1106","name":"GPT-3.5 Turbo 1106","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":1,"output":2}},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-01-14","last_updated":"2026-01-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"cohere-embed-v3-english":{"id":"cohere-embed-v3-english","type":"embedding","name":"Embed v3 English","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek-V4-Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":1.74,"output":3.48},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gpt-image-2.5-flare":{"id":"gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"cohere-embed-v-4-0":{"id":"cohere-embed-v-4-0","type":"embedding","name":"Embed v4","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":1536},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.12,"output":0}},"phi-4-reasoning-plus":{"id":"phi-4-reasoning-plus","name":"Phi-4-reasoning-plus","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25}},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-08-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"gpt-3.5-turbo-0125":{"id":"gpt-3.5-turbo-0125","name":"GPT-3.5 Turbo 0125","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":0.5,"output":1.5}},"model-router":{"id":"model-router","name":"Model Router","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"model-router","attachment":true,"reasoning":false,"tool_call":true,"release_date":"2025-05-19","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":16384},"provider":{"npm":"@ai-sdk/azure","shape":"completions"},"cost":{"input":0.14,"output":0}},"gpt-image-2.5-sunburst":{"id":"gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"phi-4-mini":{"id":"phi-4-mini","name":"Phi-4-mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"phi-4-mini-reasoning":{"id":"phi-4-mini-reasoning","name":"Phi-4-mini-reasoning","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"llama-4-maverick-17b-128e-instruct-fp8":{"id":"llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B 128E Instruct FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":16384},"provider":{"shape":"completions"},"cost":{"input":0.25,"output":1}},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-chat-latest":{"id":"gpt-chat-latest","name":"GPT Chat Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-05-05","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"beta","cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5-instant"},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.02,"output":0}},"llama-4-scout-17b-16e-instruct":{"id":"llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B 16E Instruct","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.78}},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek-V4-Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.19,"output":0.51},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"mistral-medium-2505":{"id":"mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.4,"output":2}},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-image-1":{"id":"gpt-image-1","type":"image","name":"GPT-Image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":40,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1"},"claude-mythos-5":{"id":"claude-mythos-5","name":"Claude Mythos 5","description":"Restricted Claude model for advanced cybersecurity and biology research workflows","family":"claude-mythos","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-mythos-5"},"gpt-3.5-turbo-instruct":{"id":"gpt-3.5-turbo-instruct","name":"GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-09-21","last_updated":"2023-09-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":4096},"status":"deprecated","cost":{"input":1.5,"output":2}},"phi-4-reasoning":{"id":"phi-4-reasoning","name":"Phi-4-reasoning","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"gpt-4-turbo":{"id":"gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"codestral-2501":{"id":"codestral-2501","name":"Codestral 25.01","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"provider":{"shape":"completions"},"cost":{"input":0.3,"output":0.9}},"cohere-command-a":{"id":"cohere-command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"provider":{"shape":"completions"},"cost":{"input":2.5,"output":10}},"ministral-3b":{"id":"ministral-3b","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.04,"output":0.04}},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"phi-4-multimodal":{"id":"phi-4-multimodal","name":"Phi-4-multimodal","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"phi","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.08,"output":0.32,"input_audio":4}},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}},"phi-4":{"id":"phi-4","name":"Phi-4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"status":"beta","provider":{"shape":"completions"},"canonical_model_id":"xai/grok-4.7"},"o1":{"id":"o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5}},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-07-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}}},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-4-turbo-vision":{"id":"gpt-4-turbo-vision","name":"GPT-4 Turbo Vision","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"cohere-embed-v3-multilingual":{"id":"cohere-embed-v3-multilingual","type":"embedding","name":"Embed v3 Multilingual","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"mistral-small-2503":{"id":"mistral-small-2503","name":"Mistral Small 3.1","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-03-01","last_updated":"2025-03-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.1,"output":0.3}},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-31","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"grok-4-20-reasoning":{"id":"grok-4-20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"codex-mini":{"id":"codex-mini","name":"Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-05-16","last_updated":"2025-05-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.5,"output":6,"cache_read":0.375}},"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"GPT-Image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":32,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1.5"},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"cost":{"input":0.13,"output":0}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005}},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"grok-4-20-non-reasoning":{"id":"grok-4-20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"text-embedding-ada-002":{"id":"text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.1,"output":0}},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"}}},"ebcloud":{"id":"ebcloud","env":["EBCLOUD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://maas-api.ebcloud.com/v1","name":"EBCloud","doc":"https://docs.ebtech.com/ai/model-api.html","models":{"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.4286,"output":0.8571},"canonical_model_id":"deepseek/deepseek-v4-pro"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.143,"output":0.2857},"canonical_model_id":"deepseek/deepseek-v4-flash"},"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.9286,"output":3.8571},"canonical_model_id":"moonshotai/kimi-k2.6"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.8571,"output":3.4286},"canonical_model_id":"zhipuai/glm-5.1"}}},"neuralwatt":{"id":"neuralwatt","env":["NEURALWATT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.neuralwatt.com/v1","name":"Neuralwatt","doc":"https://portal.neuralwatt.com/docs","models":{"kimi-k2.7-code-flex":{"id":"kimi-k2.7-code-flex","name":"Kimi K2.7 Code Flex","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":262128},"status":"deprecated","cost":{"input":0.6175,"output":2.6,"cache_read":0.06175},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi-k2.7-code-fast":{"id":"kimi-k2.7-code-fast","name":"Kimi K2.7 Code Fast","description":"Kimi K2.7 Code with reasoning capped to a short budget for lower latency; reasoning cannot be disabled on this model","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":262128},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.095},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"glm-5.3":{"id":"glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":1.45,"output":4.5,"cache_read":0.145},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-v4.1-flash-flex":{"id":"deepseek-v4.1-flash-flex","name":"DeepSeek V4.1 Flash Flex","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.0975,"output":0.39,"cache_read":0.00975},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"glm-5.3-flash-flex":{"id":"glm-5.3-flash-flex","name":"GLM-5.3 Flash Flex","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":0.0975,"output":0.325,"cache_read":0.0195},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen-3.8-27b-flex":{"id":"qwen-3.8-27b-flex","name":"Qwen3.8 27B Flex","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":131072},"cost":{"input":0.2925,"output":2.08,"cache_read":0.1625},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen-3.8-27b":{"id":"qwen-3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":131072},"cost":{"input":0.45,"output":3.2,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-27b"},"mimo-v2.6-pro-flex":{"id":"mimo-v2.6-pro-flex","name":"MiMo-V2.6-Pro Flex","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.5655,"output":1.131,"cache_read":0.0234},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"qwen3.6-35b-flex":{"id":"qwen3.6-35b-flex","name":"Qwen3.6 35B Flex","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131056,"output":131056},"cost":{"input":0.1885,"output":0.7475,"cache_read":0.01885},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen3.6-35b-fast":{"id":"qwen3.6-35b-fast","name":"Qwen3.6 35B Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131056,"output":131056},"cost":{"input":0.29,"output":1.15,"cache_read":0.029},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":262128},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.095},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.87,"output":1.74,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-v4-flash-speed":{"id":"deepseek-v4-flash-speed","name":"DeepSeek V4 Flash (Speed)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"kimi-k3-flex":{"id":"kimi-k3-flex","name":"Kimi K3 Flex","description":"Kimi K3 on the flex tier: discounted, best-effort latency, requests may be held under load","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":1.95,"output":9.75,"cache_read":0.195},"canonical_model_id":"moonshotai/kimi-k3"},"gemma-4-31b":{"id":"gemma-4-31b","name":"Gemma 4 31B","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":16384},"cost":{"input":0.144,"output":0.42,"cache_read":0.0144},"canonical_model_id":"google/gemma-4-31b-it"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"deepseek-v4-flash-flex":{"id":"deepseek-v4-flash-flex","name":"DeepSeek V4 Flash Flex","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.091,"output":0.182,"cache_read":0.0182},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.3-flex":{"id":"glm-5.3-flex","name":"GLM 5.3 Flex","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":0.9425,"output":2.925,"cache_read":0.09425},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.6-35b":{"id":"qwen3.6-35b","name":"Qwen3.6 35B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131056,"output":131056},"cost":{"input":0.29,"output":1.15,"cache_read":0.029},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"kimi-k3-fast":{"id":"kimi-k3-fast","name":"Kimi K3 Fast","description":"Kimi K3 with thinking disabled for low-latency tool calling, vision, and JSON work","family":"kimi-k3","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"}}},"boundless":{"id":"boundless","env":["BOUNDLESS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inference.boundless.network/v1","name":"Boundless","doc":"https://inference.boundless.network/models","models":{"qwen3.6":{"id":"qwen3.6","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.04,"output":0.35,"cache_read":0.02},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":1,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.12,"output":3.52,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.3,"output":11.4,"cache_read":0.23},"canonical_model_id":"moonshotai/kimi-k3"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01},"canonical_model_id":"openai/gpt-6-luna"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"claude-opus-5.5":{"id":"claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"anthropic/claude-opus-5-5"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.04,"output":2.08,"cache_read":0.0352},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.056},"canonical_model_id":"minimax/MiniMax-M3"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-5"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.34,"output":2.3,"cache_read":0.17},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.12,"output":0.4,"cache_read":0.024},"canonical_model_id":"zhipuai/glm-5.3-flash"},"claude-fable-5.1":{"id":"claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25},"canonical_model_id":"anthropic/claude-fable-5-1"},"dsv4":{"id":"dsv4","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.04,"output":0.21,"cache_read":0.02},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-sol"},"nemotron3-super":{"id":"nemotron3-super","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.12,"output":0.52,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"openai/gpt-6-sol"}}},"infomaniak":{"id":"infomaniak","env":["INFOMANIAK_API_KEY","INFOMANIAK_PRODUCT_ID"],"npm":"@ai-sdk/openai-compatible","api":"https://api.infomaniak.com/2/ai/${INFOMANIAK_PRODUCT_ID}/openai/v1","name":"Infomaniak","doc":"https://www.infomaniak.com/en/hosting/ai-services/open-source-models","models":{"bge_multilingual_gemma2":{"id":"bge_multilingual_gemma2","type":"embedding","name":"BGE Multilingual Gemma2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-07-25","last_updated":"2026-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"input":8000,"output":3584},"cost":{"input":0.08,"output":0}},"mini_lm_l12_v2":{"id":"mini_lm_l12_v2","type":"embedding","name":"All-MiniLM-L12-v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2021-08-30","last_updated":"2021-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128,"input":128,"output":384},"cost":{"input":0,"output":0},"canonical_model_id":"sentence-transformers/all-minilm-l12-v2"},"Qwen/Qwen3.5-122B-A10B-FP8":{"id":"Qwen/Qwen3.5-122B-A10B-FP8","name":"Qwen3.5 122B-A10B FP8","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65536},"status":"deprecated","cost":{"input":0.5,"output":3.97},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"Qwen/Qwen3.5-397B-A17B-FP8":{"id":"Qwen/Qwen3.5-397B-A17B-FP8","name":"Qwen3.5 397B-A17B FP8","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65536},"status":"beta","cost":{"input":0.99,"output":4.46},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"input":100000,"output":32768},"cost":{"input":0.25,"output":0.5},"canonical_model_id":"google/gemma-4-31b-it"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"status":"beta","cost":{"input":0.74,"output":3.72},"canonical_model_id":"moonshotai/kimi-k2.6"},"mistralai/Ministral-3-14B-Instruct-2512":{"id":"mistralai/Ministral-3-14B-Instruct-2512","name":"Ministral 3 14B Instruct","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"input":100000,"output":25600},"status":"beta","cost":{"input":0.37,"output":0.5},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"mistralai/Mistral-Small-4-119B-2603":{"id":"mistralai/Mistral-Small-4-119B-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.25,"output":0.93},"canonical_model_id":"mistral/mistral-small-2603"},"swiss-ai/Apertus-v1.5-70B":{"id":"swiss-ai/Apertus-v1.5-70B","name":"Apertus v1.5 70B","description":"Open, ethically-sourced Swiss AI model for multilingual, multimodal chat and instruction following","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-07-24","last_updated":"2026-08-01","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"input":100000,"output":8192},"status":"beta","cost":{"input":0.87,"output":3.1}},"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8":{"id":"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8","name":"Nemotron 3 Nano 30B A3B FP8","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":262144},"status":"deprecated","cost":{"input":0.06,"output":0.25},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"}}},"pareto":{"id":"pareto","env":["PARETO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.paretoinference.com/v1","name":"Pareto Inference","doc":"https://docs.paretoinference.com/","models":{"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.06,"output":0.2,"cache_read":0.012},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"ollama-cloud":{"id":"ollama-cloud","env":["OLLAMA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ollama.com/v1","name":"Ollama Cloud","doc":"https://docs.ollama.com/cloud","models":{"deepseek-v4-pro:0813":{"id":"deepseek-v4-pro:0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"nemotron-3-ultra":{"id":"nemotron-3-ultra","name":"nemotron-3-ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.1,"output":3,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"kimi-k3":{"id":"kimi-k3","name":"kimi-k3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"deepseek-v4-pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"beam":{"id":"beam","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"nemotron-3-nano:30b":{"id":"nemotron-3-nano:30b","name":"nemotron-3-nano:30b","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"kimi-k2.7-code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gemma4:31b":{"id":"gemma4:31b","name":"gemma4:31b","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-31b-it"},"gpt-oss:20b":{"id":"gpt-oss:20b","name":"gpt-oss:20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.3,"cache_read":0.035},"canonical_model_id":"openai/gpt-oss-20b"},"muse-spark-open":{"id":"muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"minimax-m2.7":{"id":"minimax-m2.7","name":"minimax-m2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":196608},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"nemotron-3-super":{"id":"nemotron-3-super","name":"nemotron-3-super","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.015,"output":0.6,"cache_read":0.015},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"kimi-k2.6":{"id":"kimi-k2.6","name":"kimi-k2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax-m3":{"id":"minimax-m3","name":"minimax-m3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":976000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"mistral-large-3:675b":{"id":"mistral-large-3:675b","name":"mistral-large-3:675b","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"deepseek-v4.1-flash:1017":{"id":"deepseek-v4.1-flash:1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"gpt-oss:120b":{"id":"gpt-oss:120b","name":"gpt-oss:120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.014},"canonical_model_id":"openai/gpt-oss-120b"}}},"moark":{"id":"moark","env":["MOARK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://moark.com/v1","name":"Moark","doc":"https://moark.com/docs/openapi/v1#tag/%E6%96%87%E6%9C%AC%E7%94%9F%E6%88%90","models":{"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.8,"output":3.6},"canonical_model_id":"deepseek/deepseek-v4-pro"},"GLM-5.3":{"id":"GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.2,"output":4.18},"canonical_model_id":"zhipuai/glm-5.3"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.5,"output":0.5},"canonical_model_id":"deepseek/deepseek-v4-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15},"canonical_model_id":"moonshotai/kimi-k3"},"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.98,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"GLM-4.7":{"id":"GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.42,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.29,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"Qwen3.6-Flash":{"id":"Qwen3.6-Flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.18,"output":1.03},"canonical_model_id":"alibaba/qwen3.6-flash"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.3},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.29,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"}}},"unorouter":{"id":"unorouter","env":["UNOROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.unorouter.com/v1","name":"UnoRouter","doc":"https://unorouter.com/models","models":{"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.2,"output":6},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.8999,"output":1.7999},"canonical_model_id":"deepseek/deepseek-v4-pro"},"minimax-m2.7:free":{"id":"minimax-m2.7:free","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"minimax/MiniMax-M2.7"},"deepseek-v4-flash:free":{"id":"deepseek-v4-flash:free","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":1.8,"output":10.8},"canonical_model_id":"openai/gpt-5.4"},"glm-4.5-flash:free":{"id":"glm-4.5-flash:free","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.819,"output":3.276},"canonical_model_id":"minimax/MiniMax-M2.7"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.0625,"output":0.125},"canonical_model_id":"deepseek/deepseek-v4-flash"},"step-3.7-flash:free":{"id":"step-3.7-flash:free","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"stepfun/step-3.7-flash"},"glm-5.2:free":{"id":"glm-5.2:free","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.5:free":{"id":"gpt-5.5:free","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.5"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.05,"output":8.4},"canonical_model_id":"openai/gpt-5.2"},"deepseek-v4-pro:free":{"id":"deepseek-v4-pro:free","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.2675,"output":5.3368},"canonical_model_id":"moonshotai/kimi-k2.6"},"qwen3.5-397b-a17b:free":{"id":"qwen3.5-397b-a17b:free","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.6001,"output":5.0288},"canonical_model_id":"zhipuai/glm-5.2"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":0.425,"output":2.125},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.44,"output":7.2},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1875,"output":1.125},"canonical_model_id":"openai/gpt-5.5"},"gpt-5.4:free":{"id":"gpt-5.4:free","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4"},"nemotron-3-ultra-550b-a55b:free":{"id":"nemotron-3-ultra-550b-a55b:free","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gemma-4-31b-it:free":{"id":"gemma-4-31b-it:free","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1857,"output":1.1142},"canonical_model_id":"google/gemini-3.5-flash"}}},"lynkr":{"id":"lynkr","env":["LYNKR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://127.0.0.1:8081/v1","name":"Lynkr","doc":"https://github.com/Fast-Editor/Lynkr","models":{"lynkr-auto":{"id":"lynkr-auto","name":"Lynkr Auto (complexity routing)","description":"Virtual model: Lynkr scores each request on complexity and routes it to the tier model the user configured (local Ollama/llama.cpp for simple requests, configured cloud providers for complex ones).","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-03","last_updated":"2026-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}}}},"xiaomi-token-plan-sgp":{"id":"xiaomi-token-plan-sgp","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan-sgp.xiaomimimo.com/v1","name":"Xiaomi Token Plan (Singapore)","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.5-tts-voicedesign":{"id":"mimo-v2.5-tts-voicedesign","type":"speech","name":"MiMo-V2.5-TTS-VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voicedesign"},"mimo-v2.5-tts-voiceclone":{"id":"mimo-v2.5-tts-voiceclone","type":"speech","name":"MiMo-V2.5-TTS-VoiceClone","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voiceclone"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo-v2.5-tts":{"id":"mimo-v2.5-tts","type":"speech","name":"MiMo-V2.5-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts"}}},"longcat":{"id":"longcat","env":["LONGCAT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.longcat.chat/openai","name":"LongCat","doc":"https://longcat.chat/platform/docs/","models":{"LongCat-2.0":{"id":"LongCat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.0"},"LongCat-2.5-Preview":{"id":"LongCat-2.5-Preview","name":"LongCat-2.5-Preview","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.5-preview"}}},"empiriolabs":{"id":"empiriolabs","env":["EMPIRIOLABS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.empiriolabs.ai/v1","name":"EmpirioLabs AI","doc":"https://docs.empiriolabs.ai","models":{"muse-spark-1-1":{"id":"muse-spark-1-1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1},"canonical_model_id":"meta/muse-spark-1.1"},"glm-4-6v-flash":{"id":"glm-4-6v-flash","name":"GLM 4.6V Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"qwen3-5-35b-a3b":{"id":"qwen3-5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.057,"output":0.459,"cache_read":0.057,"tiers":[{"input":0.229,"output":1.835,"cache_read":0.229,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"seed-2-0-code":{"id":"seed-2-0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4,"output":2.4,"cache_read":0.4,"tiers":[{"input":0.8,"output":4.8,"cache_read":0.8,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"qwen3-5-plus":{"id":"qwen3-5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":0.36,"output":2.21,"cache_read":0.36,"tiers":[{"input":1.08,"output":6.62,"cache_read":1.08,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.08,"output":6.62,"cache_read":1.08}},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen3-5-9b":{"id":"qwen3-5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":0.13,"cache_read":0.045},"canonical_model_id":"alibaba/qwen3.5-9b"},"mimo-v2-6-pro-ultraspeed":{"id":"mimo-v2-6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":4.35},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"glm-5-3-flash":{"id":"glm-5-3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.075,"output":0.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"kimi-k2-7-code-highspeed":{"id":"kimi-k2-7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":131072},"cost":{"input":1.9,"output":8,"cache_read":1.9},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":3},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16000},"cost":{"input":0.8939,"output":3.7131,"cache_read":0.1788},"canonical_model_id":"moonshotai/kimi-k2.6"},"fugu-ultra-v2-0":{"id":"fugu-ultra-v2-0","name":"Fugu Ultra v2.0","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra-v2"},"qwen3-6-35b-a3b":{"id":"qwen3-6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.07,"output":0.42,"cache_read":0.035},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.65,"output":3.3,"cache_read":1.65},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen3-5-397b-a17b":{"id":"qwen3-5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.172,"output":1.032,"cache_read":0.172,"tiers":[{"input":0.43,"output":2.58,"cache_read":0.43,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen3-6-27b":{"id":"qwen3-6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.412564,"output":2.475384,"cache_read":0.412564},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen3-5-flash":{"id":"qwen3-5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.09,"output":0.368,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.5-flash"},"fugu-max":{"id":"fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"step-3-5-flash":{"id":"step-3-5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"seed-2-0-lite":{"id":"seed-2-0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.31,"output":2.5,"cache_read":0.31,"tiers":[{"input":0.62,"output":5,"cache_read":0.62,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"qwen3-5-122b-a10b":{"id":"qwen3-5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.115,"output":0.917,"cache_read":0.115,"tiers":[{"input":0.287,"output":2.294,"cache_read":0.287,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"gemma-4-26b-a4b":{"id":"gemma-4-26b-a4b","name":"Gemma 4 26B-A4B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.29,"cache_read":0.025},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"glm-5-2":{"id":"glm-5-2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.2"},"gemma-3-27b":{"id":"gemma-3-27b","name":"Gemma 3 27B","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-3-27b-it"},"qwen3-7-max":{"id":"qwen3-7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":64000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2.5,"output":7.5,"cache_read":2.5},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen3-6-flash":{"id":"qwen3-6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":64000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.25,"tiers":[{"input":1,"output":4,"cache_read":1,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1,"output":4,"cache_read":1}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen3-8-27b":{"id":"qwen3-8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.17,"output":0.5,"cache_read":0.08},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen3-5-27b":{"id":"qwen3-5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.086,"output":0.688,"cache_read":0.086,"tiers":[{"input":0.258,"output":2.064,"cache_read":0.258,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.8,"cache_read":0.05},"canonical_model_id":"meta/muse-glimmer-30b"},"kimi-k2-7-code":{"id":"kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":131072},"cost":{"input":0.95,"output":4,"cache_read":0.95},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":1.32},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"muse-spark-1-3":{"id":"muse-spark-1-3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1},"canonical_model_id":"meta/muse-spark-1.3"},"minimax-m2-7-highspeed":{"id":"minimax-m2-7-highspeed","name":"MiniMax M2.7 Highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32768},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"qwen3-8-omni-flash":{"id":"qwen3-8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":0.94,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"mimo-v2-6-flash":{"id":"mimo-v2-6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.14},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":131072},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.7"},"muse-spark-1-2":{"id":"muse-spark-1-2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1},"canonical_model_id":"meta/muse-spark-1.2"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.424,"output":1.272,"cache_read":0.424},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"step-3-5-flash-2603":{"id":"step-3-5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0.225,"output":0.9,"cache_read":0.045,"tiers":[{"input":0.45,"output":1.8,"cache_read":0.09,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.45,"output":1.8,"cache_read":0.09}},"canonical_model_id":"minimax/MiniMax-M3"},"seed-2-1-turbo":{"id":"seed-2-1-turbo","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":0.63,"output":3.13,"cache_read":0.63},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"seed-2-0-mini":{"id":"seed-2-0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.12,"output":0.5,"cache_read":0.12,"tiers":[{"input":0.24,"output":1,"cache_read":0.24,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"deepseek-v4-1-flash-1017":{"id":"deepseek-v4-1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"qwen3-8-flash":{"id":"qwen3-8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.16},"canonical_model_id":"alibaba/qwen3.8-flash"},"mistral-small-4":{"id":"mistral-small-4","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.15},"canonical_model_id":"mistral/mistral-small-2603"},"fugu-ultra-v1-0":{"id":"fugu-ultra-v1-0","name":"Fugu Ultra v1.0","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":7.5,"output":45,"cache_read":1.5,"tiers":[{"input":15,"output":67.5,"cache_read":3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":15,"output":67.5,"cache_read":3}},"canonical_model_id":"sakana/fugu-ultra"},"seed-2-0-pro":{"id":"seed-2-0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.63,"output":3.79,"cache_read":0.63,"tiers":[{"input":1.26,"output":7.58,"cache_read":1.26,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"glm-4-7-flash":{"id":"glm-4-7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"qwen3-6-plus":{"id":"qwen3-6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.5,"tiers":[{"input":2,"output":6,"cache_read":2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":2}},"canonical_model_id":"alibaba/qwen3.6-plus"},"glm-5-1":{"id":"glm-5-1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":38912}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202000,"output":128000},"cost":{"input":0.825,"output":3.301,"cache_read":0.165,"tiers":[{"input":1.1,"output":3.851,"cache_read":0.22,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"minimax-m3-1-flash":{"id":"minimax-m3-1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.1125,"output":0.45,"cache_read":0.01125},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"qwen3-8-max":{"id":"qwen3-8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":2},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen3-8-max-0902":{"id":"qwen3-8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":2},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen3-7-plus":{"id":"qwen3-7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":256000}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.4,"tiers":[{"input":1.2,"output":4.8,"cache_read":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"step-3-7-flash":{"id":"step-3-7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":131072},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"glm-5-3":{"id":"glm-5-3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.3"},"mimo-v2-6-pro":{"id":"mimo-v2-6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.435},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"fugu-ultra-v1-1":{"id":"fugu-ultra-v1-1","name":"Fugu Ultra v1.1","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"glm-4-5-flash":{"id":"glm-4-5-flash","name":"GLM 4.5 Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"qwen3-7-flash":{"id":"qwen3-7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen3-5-4b":{"id":"qwen3-5-4b","name":"Qwen3.5 4B","description":"Qwen3.5 4B is a low-cost multimodal reasoning model with 256K context, image and video input, function tools, and structured output.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.04,"output":0.07,"cache_read":0.02},"canonical_model_id":"alibaba/qwen3.5-4b"}}},"klokintegration":{"id":"klokintegration","env":["KLOKINTEGRATION_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api-gw.klok.ipaas.se/proxy/kloker-key/v1","name":"klokintegration.se","doc":"https://klokintegration.se/docs/ai-api","models":{"Kloker-Integration-Architect":{"id":"Kloker-Integration-Architect","name":"Kloker Integration Architect","description":"Knows the customer integration environment and Klok best practices. Opinionated about structure. The gateway runs ecosystem lookup tools server-side and appends a system-prompt injection (data contracts, CloudEvents, event-driven flows). Client system prompts and OpenAI tool calls are preserved.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":50000},"status":"beta","cost":{"input":0.23,"output":1.16}},"Kloker":{"id":"Kloker","name":"Kloker","description":"Cheap general model with a clean context. Nothing from the customer environment is packed in. It tracks the current best open source model. The Klok team verifies it and upgrades it periodically.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":50000},"status":"beta","cost":{"input":0.23,"output":1.16}},"Kloker-Integration-Developer":{"id":"Kloker-Integration-Developer","name":"Kloker Integration Developer","description":"Knows the customer integration environment and Klok best practices. Opinionated about implementation. The gateway runs ecosystem lookup tools server-side and appends a system-prompt injection. Client system prompts and OpenAI tool calls are preserved.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":50000},"status":"beta","cost":{"input":0.23,"output":1.16}}}},"venice":{"id":"venice","env":["VENICE_API_KEY"],"npm":"venice-ai-sdk-provider","name":"Venice AI","doc":"https://docs.venice.ai","models":{"openai-gpt-6-sol":{"id":"openai-gpt-6-sol","name":"GPT-6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"llama-3.2-3b":{"id":"llama-3.2-3b","name":"Llama 3.2 3B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"meta/llama-3.2-3b"},"zai-org-glm-4.7":{"id":"zai-org-glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":16384},"cost":{"input":0.55,"output":2.65,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"qwen3-5-35b-a3b":{"id":"qwen3-5-35b-a3b","name":"Qwen 3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.15,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen-3-8-27b":{"id":"qwen-3-8-27b","name":"Qwen 3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.45,"output":3.2},"canonical_model_id":"alibaba/qwen3.8-27b"},"openai-gpt-oss-120b":{"id":"openai-gpt-oss-120b","name":"OpenAI GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-120b"},"gemini-3-7-flash":{"id":"gemini-3-7-flash","name":"Gemini 3.7 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.9375,"output":4.6875,"cache_read":0.09375},"canonical_model_id":"google/gemini-3.7-flash"},"kimi-k3-1-pro-highspeed":{"id":"kimi-k3-1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":6,"output":30,"cache_read":0.6000000000000001}},"olafangensan-glm-4.7-flash-heretic":{"id":"olafangensan-glm-4.7-flash-heretic","name":"GLM 4.7 Flash Heretic","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-02-04","last_updated":"2026-06-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":24000},"cost":{"input":0.07,"output":0.4,"cache_read":0.035}},"z-ai-glm-5-3":{"id":"z-ai-glm-5-3","name":"GLM 5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.75,"output":5.5,"cache_read":0.325},"canonical_model_id":"zhipuai/glm-5.3"},"xiaomi-mimo-v2-5":{"id":"xiaomi-mimo-v2-5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2,"cache_read":0.08},"canonical_model_id":"xiaomi/mimo-v2.5"},"hermes-3-llama-3.1-405b":{"id":"hermes-3-llama-3.1-405b","name":"Hermes 3 Llama 3.1 405b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":1.1,"output":3},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"qwen-3-6-plus":{"id":"qwen-3-6-plus","name":"Qwen 3.6 Plus Uncensored","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.625,"output":3.75,"cache_read":0.0625,"cache_write":0.78,"tiers":[{"input":2.5,"output":7.5,"cache_read":0.0625,"cache_write":0.78,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2.5,"output":7.5,"cache_read":0.0625,"cache_write":0.78}},"canonical_model_id":"alibaba/qwen3.6-plus"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3.75,"output":18.75,"cache_read":0.375},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.75,"output":3.5,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"openai-gpt-52":{"id":"openai-gpt-52","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":272000,"output":65536},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"qwen3-6-35b-a3b":{"id":"qwen3-6-35b-a3b","name":"Qwen 3.6 35B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.1,"output":1},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"zai-org-glm-4.6":{"id":"zai-org-glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":16384},"cost":{"input":0.43,"output":1.75,"cache_read":0.08},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai-glm-5-3-flash":{"id":"z-ai-glm-5-3-flash","name":"GLM 5.3 Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen-3-8-2-4t-a95b":{"id":"qwen-3-8-2-4t-a95b","name":"Qwen 3.8 2.4T","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":2.5,"output":7.5,"cache_read":0.3125},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":1.65,"output":3.301,"cache_read":0.33},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen3-5-397b-a17b":{"id":"qwen3-5-397b-a17b","name":"Qwen 3.5 397B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.75,"output":4.5},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen3-6-27b":{"id":"qwen3-6-27b","name":"Qwen 3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.325,"output":3.25},"canonical_model_id":"alibaba/qwen3.6-27b"},"openai-gpt-6-luna":{"id":"openai-gpt-6-luna","name":"GPT-6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai-gpt-53-codex":{"id":"openai-gpt-53-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai-gpt-61-sol":{"id":"openai-gpt-61-sol","name":"GPT-6.1 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"qwen3-next-80b":{"id":"qwen3-next-80b","name":"Qwen 3 Next 80b","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.35,"output":1.9},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"deepseek-v4-flash-0731-fast":{"id":"deepseek-v4-flash-0731-fast","name":"DeepSeek V4 Flash 0731 Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.35,"output":0.7,"cache_read":0.0875},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"Qwen3 VL 235B","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-01-16","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.21,"output":1.9,"cache_read":0.1}},"openai-gpt-4o-2024-11-20":{"id":"openai-gpt-4o-2024-11-20","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":198000,"output":32768},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"abliteration-abliterated-model-large-v2":{"id":"abliteration-abliterated-model-large-v2","name":"Abliterated Large V2","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":3,"output":5,"cache_read":0.3}},"claude-opus-5-5-fast":{"id":"claude-opus-5-5-fast","name":"Claude Opus 5.5 Fast","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":8,"output":40,"cache_read":0.4,"cache_write":10},"canonical_model_id":"anthropic/claude-opus-5-5"},"gemini-3-5-flash-lite":{"id":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.375,"output":3.125,"cache_read":0.0375},"canonical_model_id":"google/gemini-3.5-flash-lite"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen 3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.75},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"qwen-3-8-max":{"id":"qwen-3-8-max","name":"Qwen 3.8 Max","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.3125,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"mercury-2-5":{"id":"mercury-2-5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.25,"output":0.9375,"cache_read":0.025},"canonical_model_id":"inception/mercury-2.5"},"zai-org-glm-5":{"id":"zai-org-glm-5","name":"GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":32000},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"openai-gpt-56-sol-pro":{"id":"openai-gpt-56-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"z-ai-glm-5-turbo":{"id":"z-ai-glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32768},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"mercury-3":{"id":"mercury-3","name":"Mercury 3","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01},"canonical_model_id":"inception/mercury-3"},"gemini-3-6-flash":{"id":"gemini-3-6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.9375,"output":4.6875,"cache_read":0.09375},"canonical_model_id":"google/gemini-3.6-flash"},"openai-gpt-56-sol":{"id":"openai-gpt-56-sol","name":"GPT-5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"minimax-m27":{"id":"minimax-m27","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":32768},"cost":{"input":0.375,"output":1.5,"cache_read":0.06875},"canonical_model_id":"minimax/MiniMax-M2.7"},"muse-spark-open":{"id":"muse-spark-open","name":"Muse Spark Open","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.5625,"output":5.3125,"cache_read":0.1875},"canonical_model_id":"meta/muse-spark-open"},"aion-labs-aion-3-0":{"id":"aion-labs-aion-3-0","name":"Aion 3.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":3.75,"output":7.5,"cache_read":0.9375},"canonical_model_id":"aion-labs/aion-3.0"},"nvidia-nemotron-3-nano-30b-a3b":{"id":"nvidia-nemotron-3-nano-30b-a3b","name":"NVIDIA Nemotron 3 Nano 30B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.075,"output":0.3},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"qwen-3-7-plus":{"id":"qwen-3-7-plus","name":"Qwen 3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":2,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":1.5,"output":6,"cache_read":0.15,"cache_write":1.875,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.5,"output":6,"cache_read":0.15,"cache_write":1.875}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen3-235b-a22b-thinking-2507":{"id":"qwen3-235b-a22b-thinking-2507","name":"Qwen 3 235B A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.45,"output":3.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"venice-uncensored-1-2":{"id":"venice-uncensored-1-2","name":"Venice Uncensored 1.2","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"venice","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-04-01","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2,"output":0.9}},"minimax-m3-preview":{"id":"minimax-m3-preview","name":"MiniMax M3 Preview","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax-m3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"release_date":"2026-06-12","last_updated":"2026-06-13","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.3,"output":1.2,"cache_read":0.06}},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash 0423","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.138,"output":0.275,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"openai-gpt-61-sol-ultrafast":{"id":"openai-gpt-61-sol-ultrafast","name":"GPT-6.1 Sol Ultrafast","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":12,"output":60,"cache_read":0.6,"cache_write":15,"tiers":[{"input":24,"output":90,"cache_read":1.2,"cache_write":30,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":24,"output":90,"cache_read":1.2,"cache_write":30}},"canonical_model_id":"openai/gpt-6.1-sol"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2.5,"output":12.5,"cache_read":0.125},"canonical_model_id":"google/gemini-4-argon"},"kimi-k2-7-code":{"id":"kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.75,"output":3.5,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"google-gemma-4-26b-a4b-it":{"id":"google-gemma-4-26b-a4b-it","name":"Google Gemma 4 26B A4B Instruct","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.13,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":1.65,"output":4.95,"cache_read":0.165},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"gemma-4-uncensored":{"id":"gemma-4-uncensored","name":"Gemma 4 Uncensored","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-04-13","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.1625,"output":0.5}},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"qwen3-coder-480b-a35b-instruct-turbo":{"id":"qwen3-coder-480b-a35b-instruct-turbo","name":"Qwen 3 Coder 480B Turbo","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-01-27","last_updated":"2026-06-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.35,"output":1.5,"cache_read":0.04}},"openai-gpt-56-terra-pro":{"id":"openai-gpt-56-terra-pro","name":"GPT-5.6 Terra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.0075},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"openai-gpt-55":{"id":"openai-gpt-55","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"llama-3.3-70b":{"id":"llama-3.3-70b","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.7,"output":2.8},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"grok-4-20-multi-agent":{"id":"grok-4-20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"release_date":"2026-03-12","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":1.42,"output":2.83,"cache_read":0.23,"tiers":[{"input":2.83,"output":5.67,"cache_read":0.45,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.83,"output":5.67,"cache_read":0.45}}},"kimi-k3-1-pro":{"id":"kimi-k3-1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.30000000000000004},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"aion-labs-aion-3-0-mini":{"id":"aion-labs-aion-3-0-mini","name":"Aion 3.0 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":0.875,"output":1.75,"cache_read":0.225},"canonical_model_id":"aion-labs/aion-3.0-mini"},"gemini-3-1-pro-preview":{"id":"gemini-3-1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2.5,"output":15,"cache_read":0.5,"cache_write":0.5,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"cache_write":0.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5,"cache_write":0.5}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"openai-gpt-56-luna-pro":{"id":"openai-gpt-56-luna-pro","name":"GPT-5.6 Luna Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":198000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"openai-gpt-6-astra-pro":{"id":"openai-gpt-6-astra-pro","name":"GPT-6 Astra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"qwen-3-7-max":{"id":"qwen-3-7-max","name":"Qwen 3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2.7,"output":8.05,"cache_read":0.27,"cache_write":3.35},"canonical_model_id":"alibaba/qwen3.7-max"},"grok-4-7":{"id":"grok-4-7","name":"Grok 4.7","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":200000},"cost":{"input":2.27,"output":6.8,"cache_read":0.57,"tiers":[{"input":4.53,"output":13.6,"cache_read":1.13,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.53,"output":13.6,"cache_read":1.13}},"canonical_model_id":"xai/grok-4.7"},"openai-gpt-4o-mini-2024-07-18":{"id":"openai-gpt-4o-mini-2024-07-18","name":"GPT-4o Mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"venice-uncensored-role-play":{"id":"venice-uncensored-role-play","name":"Venice Role Play Uncensored","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"venice","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-02-20","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.5,"output":2}},"claude-opus-5-fast":{"id":"claude-opus-5-fast","name":"Claude Opus 5 Fast","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-5"},"mercury-2":{"id":"mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.3125,"output":0.9375,"cache_read":0.03125},"canonical_model_id":"inception/mercury-2"},"nvidia-nemotron-3-5-super-vl-120b-a12b":{"id":"nvidia-nemotron-3-5-super-vl-120b-a12b","name":"NVIDIA Nemotron 3.5 Super VL 120B","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.25,"output":1.25,"cache_read":0.125},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"grok-4-20":{"id":"grok-4-20","name":"Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-03-12","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":1.42,"output":2.83,"cache_read":0.23,"tiers":[{"input":2.83,"output":5.67,"cache_read":0.45,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.83,"output":5.67,"cache_read":0.45}}},"gemini-3-8-flash":{"id":"gemini-3-8-flash","name":"Gemini 3.8 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.9375,"output":4.6875,"cache_read":0.09375},"canonical_model_id":"google/gemini-3.8-flash"},"seed-2-1-turbo":{"id":"seed-2-1-turbo","name":"Seed 2.1 Turbo","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":0.625,"output":3.125,"cache_read":0.125},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"deepseek-v4-1-flash-1017":{"id":"deepseek-v4-1-flash-1017","name":"DeepSeek v4.1 Flash (1017)","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.0075},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"qwen-3-8-flash":{"id":"qwen-3-8-flash","name":"Qwen 3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.14,"output":0.49,"cache_read":0.014},"canonical_model_id":"alibaba/qwen3.8-flash"},"openai-gpt-54":{"id":"openai-gpt-54","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":131072},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"mistral-small-3-2-24b-instruct":{"id":"mistral-small-3-2-24b-instruct","name":"Mistral Small 3.2 24B Instruct","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.09375,"output":0.25},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"minimax-m25":{"id":"minimax-m25","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":32768},"cost":{"input":0.27,"output":0.95,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":0.7,"output":3.75,"cache_read":0.07},"canonical_model_id":"google/gemini-3-flash-preview"},"aion-labs-aion-3-5-mini":{"id":"aion-labs-aion-3-5-mini","name":"Aion 3.5 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.875,"output":1.75,"cache_read":0.225}},"google-gemma-3-27b-it":{"id":"google-gemma-3-27b-it","name":"Google Gemma 3 27B Instruct","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":16384},"cost":{"input":0.12,"output":0.2},"canonical_model_id":"google/gemma-3-27b-it"},"openai-gpt-56-terra":{"id":"openai-gpt-56-terra","name":"GPT-5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"xiaomi-mimo-v2-6-flash":{"id":"xiaomi-mimo-v2-6-flash","name":"MiMo-V2.6-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.175,"output":0.35,"cache_read":0.00375},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"grok-4-3":{"id":"grok-4-3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":1.42,"output":2.83,"cache_read":0.23,"tiers":[{"input":2.83,"output":5.67,"cache_read":0.45,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.83,"output":5.67,"cache_read":0.45}},"canonical_model_id":"xai/grok-4.3"},"google-gemma-4-31b-it":{"id":"google-gemma-4-31b-it","name":"Google Gemma 4 31B Instruct","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.12,"output":0.36,"cache_read":0.09},"canonical_model_id":"google/gemma-4-31b-it"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"openai-gpt-6-astra":{"id":"openai-gpt-6-astra","name":"GPT-6 Astra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"status":"deprecated","cost":{"input":0.5625,"output":1.5,"cache_read":0.125},"canonical_model_id":"thinkingmachines/inkling"},"muse-wonder":{"id":"muse-wonder","name":"Muse Wonder","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-12","last_updated":"2026-10-15","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":5,"output":25,"cache_read":0.25}},"kimi-k2-5":{"id":"kimi-k2-5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.56,"output":3.5,"cache_read":0.22},"canonical_model_id":"moonshotai/kimi-k2.5"},"zai-org-glm-5-1":{"id":"zai-org-glm-5-1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":80000},"cost":{"input":1.54,"output":4.84,"cache_read":0.286},"canonical_model_id":"zhipuai/glm-5.1"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 Preview","description":"Preview model for early access evaluation, prototyping, and compatibility testing","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":1.0425,"output":3.12625,"cache_read":0.0525},"canonical_model_id":"tencent/hy4-preview"},"abliteration-abliterated-model-large-flash":{"id":"abliteration-abliterated-model-large-flash","name":"Abliterated Large Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.125}},"openai-gpt-56-luna":{"id":"openai-gpt-56-luna","name":"GPT-5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai-gpt-55-pro":{"id":"openai-gpt-55-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"zai-org-glm-5-2":{"id":"zai-org-glm-5-2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4-1-flash-uncensored":{"id":"deepseek-v4-1-flash-uncensored","name":"DeepSeek V4.1 Flash Uncensored","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.33,"output":1.32,"cache_read":0.00825}},"minimax-m3-1-flash":{"id":"minimax-m3-1-flash","name":"MiniMax M3.1 Flash","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1875,"output":0.75,"cache_read":0.01875},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":200000},"cost":{"input":2.27,"output":6.8,"cache_read":0.57,"tiers":[{"input":4.53,"output":13.6,"cache_read":1.13,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.53,"output":13.6,"cache_read":1.13}},"canonical_model_id":"xai/grok-4.6"},"grok-build-0-1":{"id":"grok-build-0-1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"openai-gpt-54-pro":{"id":"openai-gpt-54-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"kimi-k3-fast-api":{"id":"kimi-k3-fast-api","name":"Kimi K3 Fast","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"grok-4-5":{"id":"grok-4-5","name":"Grok 4.5","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":32000},"cost":{"input":2.27,"output":6.8,"cache_read":0.34,"tiers":[{"input":4.53,"output":13.6,"cache_read":0.68,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.53,"output":13.6,"cache_read":0.68}},"canonical_model_id":"xai/grok-4.5"},"nvidia-nemotron-3-ultra-550b-a55b":{"id":"nvidia-nemotron-3-ultra-550b-a55b","name":"NVIDIA Nemotron 3 Ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0.625,"output":3.125,"cache_read":0.1875},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"openai-gpt-54-mini":{"id":"openai-gpt-54-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"aion-labs-aion-3-5":{"id":"aion-labs-aion-3-5","name":"Aion 3.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":3.75,"output":7.5,"cache_read":0.9375}},"z-ai-glm-5v-turbo":{"id":"z-ai-glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32768},"cost":{"input":1.5,"output":5,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5v-turbo"},"gemini-3-5-flash":{"id":"gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.55,"output":9.45,"cache_read":0.155,"cache_write":0.086},"canonical_model_id":"google/gemini-3.5-flash"},"claude-opus-4-8-fast":{"id":"claude-opus-4-8-fast","name":"Claude Opus 4.8 Fast","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":160000,"output":32768},"status":"deprecated","cost":{"input":0.33,"output":0.48,"cache_read":0.16},"canonical_model_id":"deepseek/deepseek-v3.2"}}},"drun":{"id":"drun","env":["DRUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://chat.d.run/v1","name":"D.Run (China)","doc":"https://www.d.run","models":{"public/deepseek-v3":{"id":"public/deepseek-v3","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.28,"output":1.1},"canonical_model_id":"deepseek/deepseek-v3"},"public/minimax-m25":{"id":"public/minimax-m25","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.29,"output":1.16},"canonical_model_id":"minimax/MiniMax-M2.5"},"public/deepseek-r1":{"id":"public/deepseek-r1","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"deepseek/deepseek-r1"}}},"subconscious":{"id":"subconscious","env":["SUBCONSCIOUS_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.subconscious.dev/v1","name":"Subconscious","doc":"https://docs.subconscious.dev","models":{"subconscious/deepseek-v4.1-flash-marathon":{"id":"subconscious/deepseek-v4.1-flash-marathon","name":"DeepSeek V4.1 Flash Marathon","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":5000000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"subconscious/glm-5.3-marathon":{"id":"subconscious/glm-5.3-marathon","name":"GLM-5.3 Marathon","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":5000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"}}},"deepinfra":{"id":"deepinfra","env":["DEEPINFRA_API_KEY"],"npm":"@ai-sdk/deepinfra","name":"Deep Infra","doc":"https://deepinfra.com/models","models":{"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"Qwen/Qwen3.8-Max":{"id":"Qwen/Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":1.65,"output":4.951,"cache_read":0.206},"canonical_model_id":"alibaba/qwen3.8-max"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.32,"output":3.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo","name":"Qwen3 Coder 480B A35B Instruct Turbo","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":66536},"cost":{"input":0.3,"output":1,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen 3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26,"output":2.6},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-30B-A3B":{"id":"Qwen/Qwen3-30B-A3B","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"Qwen/Qwen3-VL-235B-A22B-Instruct":{"id":"Qwen/Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"Qwen/Qwen3.7-Max":{"id":"Qwen/Qwen3.7-Max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"tiers":[{"input":5,"output":15,"cache_read":1,"tier":{"type":"context","size":32000}},{"input":6.25,"output":18.5,"cache_read":1.25,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.7-max"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.1,"output":0.95,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.8-Flash":{"id":"Qwen/Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.113,"output":0.382,"cache_read":0.0141},"canonical_model_id":"alibaba/qwen3.8-flash"},"Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":2.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3-Max":{"id":"Qwen/Qwen3-Max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"status":"deprecated","cost":{"input":1.2,"output":6,"cache_read":0.24,"tiers":[{"input":2.4,"output":12,"cache_read":0.48,"tier":{"type":"context","size":32000}},{"input":3,"output":15,"cache_read":0.6,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen 3.5 397B A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.45,"output":3,"cache_read":0.22},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.29,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.25,"output":1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.056},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.15,"output":1.15,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"status":"deprecated","cost":{"input":0.95,"output":4.05,"cache_read":0.16},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"ByteDance/Seed-2.0-pro":{"id":"ByteDance/Seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.5,"output":3,"cache_read":0.1,"tiers":[{"input":1,"output":6,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"ByteDance/Seed-2.0-mini":{"id":"ByteDance/Seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02,"tiers":[{"input":0.2,"output":0.8,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"ByteDance/Seed-2.0-code":{"id":"ByteDance/Seed-2.0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1,"tiers":[{"input":1,"output":6,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemma-4-E4B-it":{"id":"google/gemma-4-E4B-it","name":"Gemma 4 E4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.02,"output":0.1},"canonical_model_id":"google/gemma-4-E4B-it"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.03,"output":0.14},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.037,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.06,"output":0.18,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"cost":{"input":0.09,"output":0.18,"cache_read":0.018,"cache_write":0.1125},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":8192},"cost":{"input":0.32,"output":0.89},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"id":"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":8192},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek-R1-0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":64000},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":64000},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"status":"deprecated","cost":{"input":0.24,"output":0.9,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3-0324"},"stepfun-ai/Step-3.7-Flash":{"id":"stepfun-ai/Step-3.7-Flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"status":"deprecated","cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun-ai/Step-5-Preview":{"id":"stepfun-ai/Step-5-Preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.75,"output":3.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.85,"output":14.25,"cache_read":0.285,"cache_write":3.5625},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.68,"output":3.4,"cache_read":0.136,"cache_write":0.85},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"mistralai/Mistral-Small-24B-Instruct-2501":{"id":"mistralai/Mistral-Small-24B-Instruct-2501","name":"Mistral Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"tencent/Hy4-preview":{"id":"tencent/Hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/Hy3":{"id":"tencent/Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"status":"deprecated","cost":{"input":0.13,"output":0.53,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"id":"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","name":"Llama 4 Maverick 17B FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":0.8},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/Llama-3.3-70B-Instruct-Turbo":{"id":"meta-llama/Llama-3.3-70B-Instruct-Turbo","name":"Llama 3.3 70B Turbo","description":"Compact Llama instruction model for fast chat and local deployment","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/Llama-Guard-4-12B":{"id":"meta-llama/Llama-Guard-4-12B","name":"Llama Guard 4 12B","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"meta-llama/Llama-4-Scout-17B-16E-Instruct":{"id":"meta-llama/Llama-4-Scout-17B-16E-Instruct","name":"Llama 4 Scout 17B","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning":{"id":"nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.2,"output":0.8},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/Nemotron-3-Nano-30B-A3B":{"id":"nvidia/Nemotron-3-Nano-30B-A3B","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.05,"output":0.2,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/Llama-3.3-Nemotron-Super-49B-v1.5":{"id":"nvidia/Llama-3.3-Nemotron-Super-49B-v1.5","name":"Llama 3.3 Nemotron Super 49B v1.5","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":0.4},"canonical_model_id":"nvidia/llama-3.3-nemotron-super-49b-v1.5"},"XiaomiMiMo/MiMo-V2.5":{"id":"XiaomiMiMo/MiMo-V2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"XiaomiMiMo/MiMo-V2.6-Flash":{"id":"XiaomiMiMo/MiMo-V2.6-Flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"XiaomiMiMo/MiMo-V2.5-Pro":{"id":"XiaomiMiMo/MiMo-V2.5-Pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"status":"deprecated","cost":{"input":1,"output":3,"cache_read":0.2},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"XiaomiMiMo/MiMo-V2.6-Pro":{"id":"XiaomiMiMo/MiMo-V2.6-Pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.9,"output":4,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"cost":{"input":0.4,"output":1.75,"cache_read":0.08},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":1.05,"output":3.5,"cache_read":0.205},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-4.7-Flash":{"id":"zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0.6,"output":2.08,"cache_read":0.12},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.75,"output":2.4,"cache_read":0.14,"cache_write":0.9375},"canonical_model_id":"zhipuai/glm-5.2"}}},"crusoe":{"id":"crusoe","env":["CRUSOE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inference.crusoecloud.com/v1","name":"Crusoe","doc":"https://docs.crusoecloud.com/managed-inference/overview","models":{"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.22,"output":0.8,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"zai/GLM-5.1":{"id":"zai/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4.4,"cache_read":0.25},"canonical_model_id":"zhipuai/glm-5.1"},"zai/GLM-5.2":{"id":"zai/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4,"cache_read":0.14},"canonical_model_id":"google/gemma-4-31b-it"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.05},"canonical_model_id":"openai/gpt-oss-120b"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"status":"deprecated","cost":{"input":0.5,"output":1.5,"cache_read":0.25},"canonical_model_id":"deepseek/deepseek-v3-0324"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.7,"output":3.5,"cache_read":0.35},"canonical_model_id":"moonshotai/kimi-k2.6"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.25,"output":0.75,"cache_read":0.13},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B":{"id":"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.03},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B":{"id":"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.4,"cache_read":0.15},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/Nemotron-3-Nano-Omni-Reasoning-30B-A3B":{"id":"nvidia/Nemotron-3-Nano-Omni-Reasoning-30B-A3B","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.3,"output":1.83,"cache_read":0.3,"input_audio":0.5},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"}}},"minimax-coding-plan":{"id":"minimax-coding-plan","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.io/anthropic/v1","name":"MiniMax Token Plan (minimax.io)","doc":"https://platform.minimax.io/docs/token-plan/intro","models":{"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash-Preview":{"id":"MiniMax-M3.1-Flash-Preview","name":"MiniMax-M3.1-Flash-Preview","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-27","last_updated":"2026-09-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash-Preview"},"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"}}},"dinference":{"id":"dinference","env":["DINFERENCE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.dinference.com/v1","name":"DInference","doc":"https://dinference.com","models":{"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.75,"output":2.4},"canonical_model_id":"zhipuai/glm-5"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":0.22,"output":0.88},"canonical_model_id":"minimax/MiniMax-M2.5"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":3.89},"canonical_model_id":"zhipuai/glm-5.1"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.45,"output":1.65},"canonical_model_id":"zhipuai/glm-4.7"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":3.89},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.0675,"output":0.27},"canonical_model_id":"openai/gpt-oss-120b"}}},"fireworks-ai":{"id":"fireworks-ai","env":["FIREWORKS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.fireworks.ai/inference/v1/","name":"Fireworks AI","doc":"https://fireworks.ai/docs/","models":{"accounts/fireworks/models/kimi-k3":{"id":"accounts/fireworks/models/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3.75,"output":18.75,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/models/deepseek-v4p1-flash-1017":{"id":"accounts/fireworks/models/deepseek-v4p1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"experimental":{"modes":{"priority":{"cost":{"input":0.375,"output":1.5,"cache_read":0.0075},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"accounts/fireworks/models/muse-spark-open":{"id":"accounts/fireworks/models/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"accounts/fireworks/models/minimax-m3p1-flash":{"id":"accounts/fireworks/models/minimax-m3p1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"accounts/fireworks/models/qwen3p8-max":{"id":"accounts/fireworks/models/qwen3p8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3,"output":9,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"accounts/fireworks/models/nemotron-3-ultra-nvfp4":{"id":"accounts/fireworks/models/nemotron-3-ultra-nvfp4","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"accounts/fireworks/models/minimax-m3":{"id":"accounts/fireworks/models/minimax-m3","name":"MiniMax-M3","description":"Fireworks text-only MiniMax coding model for long-context reasoning and agent tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"experimental":{"modes":{"priority":{"cost":{"input":0.45,"output":1.8,"cache_read":0.09},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"accounts/fireworks/models/nemotron-lightning-3p5-30b-a3b":{"id":"accounts/fireworks/models/nemotron-lightning-3p5-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","experimental":{"modes":{"priority":{"cost":{"input":0.0625,"output":0.25,"cache_read":0.0125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"accounts/fireworks/models/inkling":{"id":"accounts/fireworks/models/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"status":"deprecated","cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"accounts/fireworks/models/gpt-oss-120b":{"id":"accounts/fireworks/models/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","experimental":{"modes":{"priority":{"cost":{"input":0.18,"output":0.72,"cache_read":0.018},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"},"accounts/fireworks/models/glm-5p3":{"id":"accounts/fireworks/models/glm-5p3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":262144},"experimental":{"modes":{"priority":{"cost":{"input":1.75,"output":5.5,"cache_read":0.325},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/models/glm-5p3-flash":{"id":"accounts/fireworks/models/glm-5p3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":0.1875,"output":0.625,"cache_read":0.0375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"accounts/fireworks/models/qwen3p8-2p4t-a95b":{"id":"accounts/fireworks/models/qwen3p8-2p4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"accounts/fireworks/models/ember-1":{"id":"accounts/fireworks/models/ember-1","name":"Ember-1","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"accounts/fireworks/models/deepseek-v4p1-flash":{"id":"accounts/fireworks/models/deepseek-v4p1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"experimental":{"modes":{"priority":{"cost":{"input":0.375,"output":1.5,"cache_read":0.0075},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"accounts/fireworks/routers/glm-5p3-fast":{"id":"accounts/fireworks/routers/glm-5p3-fast","name":"GLM 5.3 Fast","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048572,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.39},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/routers/glm-flash-latest":{"id":"accounts/fireworks/routers/glm-flash-latest","name":"GLM Flash Latest (GLM 5.3 Flash)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":0.1875,"output":0.625,"cache_read":0.0375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"accounts/fireworks/routers/glm-5p3-spot":{"id":"accounts/fireworks/routers/glm-5p3-spot","name":"GLM 5.3 Spot","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":262144},"cost":{"input":0.98,"output":3.08,"cache_read":0.182},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/routers/glm-fast-latest":{"id":"accounts/fireworks/routers/glm-fast-latest","name":"GLM 5.3 Fast (Latest)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048572,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.39},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/routers/deepseek-v4p1-flash-1017-spot":{"id":"accounts/fireworks/routers/deepseek-v4p1-flash-1017-spot","name":"DeepSeek V4.1 Flash 1017 Spot","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.21,"output":0.84,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"accounts/fireworks/routers/glm-latest":{"id":"accounts/fireworks/routers/glm-latest","name":"GLM Latest","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":262144},"experimental":{"modes":{"priority":{"cost":{"input":1.75,"output":5.5,"cache_read":0.325},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/routers/kimi-fast-latest":{"id":"accounts/fireworks/routers/kimi-fast-latest","name":"Kimi Fast Latest","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/minimax-latest":{"id":"accounts/fireworks/routers/minimax-latest","name":"MiniMax Latest","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"experimental":{"modes":{"priority":{"cost":{"input":0.45,"output":1.8,"cache_read":0.09},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"accounts/fireworks/routers/qwen-max-latest":{"id":"accounts/fireworks/routers/qwen-max-latest","name":"Qwen Max Latest (Qwen3.8 Max)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3,"output":9,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"accounts/fireworks/routers/deepseek-flash-latest":{"id":"accounts/fireworks/routers/deepseek-flash-latest","name":"DeepSeek Flash Latest","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"experimental":{"modes":{"priority":{"cost":{"input":0.375,"output":1.5,"cache_read":0.0075},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"accounts/fireworks/routers/kimi-k3-spot":{"id":"accounts/fireworks/routers/kimi-k3-spot","name":"Kimi K3 Spot","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.1,"output":10.5,"cache_read":0.21},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/kimi-latest":{"id":"accounts/fireworks/routers/kimi-latest","name":"Kimi Latest","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3.75,"output":18.75,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/kimi-k3-fast":{"id":"accounts/fireworks/routers/kimi-k3-fast","name":"Kimi K3 Fast","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/deepseek-v4p1-flash-spot":{"id":"accounts/fireworks/routers/deepseek-v4p1-flash-spot","name":"DeepSeek V4.1 Flash Spot","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.21,"output":0.84,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4.1-flash"}}},"kimi-code-plan-cn":{"id":"kimi-code-plan-cn","env":["KIMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.kimi.com/coding/v1","name":"Kimi For Coding (kimi.com)","doc":"https://www.kimi.com/code/docs/en/kimi-code/models.html","models":{"k3":{"id":"k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"k3.1-pro":{"id":"k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-for-coding":{"id":"kimi-for-coding","name":"Kimi K2.8 Preview","description":"Kimi coding model with more efficient thinking and up to 1M context, available through Kimi Code","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.8-preview"},"k3-256k":{"id":"k3-256k","name":"Kimi K3-256K","description":"256K-context version of Kimi K3, reducing token consumption for shorter coding sessions","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-for-coding-highspeed":{"id":"kimi-for-coding-highspeed","name":"Kimi For Coding HighSpeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"k3.1-pro-highspeed":{"id":"k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"}}},"aixy":{"id":"aixy","env":["AIXY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.aixy-gateway.com/v1","name":"Aixy","doc":"https://docs.aixy-gateway.com/integrations/overview","models":{"gemini/gemini-4-argon":{"id":"gemini/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"gemini/gemini-3.8-flash":{"id":"gemini/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"}}}} \ No newline at end of file +{"aimlapi":{"id":"aimlapi","env":["AIMLAPI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.aimlapi.com/v1","name":"AI/ML API","doc":"https://docs.aimlapi.com","models":{"mistralai/mistral-large-2407":{"id":"mistralai/mistral-large-2407","name":"Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2.116,"output":6.348,"cache_read":0.2116},"canonical_model_id":"mistral/mistral-large-2407"},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.3703,"output":0.3703,"cache_read":0.02116},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2 2512","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.46552,"output":2.3276,"cache_read":0.046552},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2.116,"output":6.348,"cache_read":0.2116},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4232,"output":2.116,"cache_read":0.04232},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 2508","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":0.3174,"output":0.9522,"cache_read":0.03174},"canonical_model_id":"mistral/codestral-2508"},"mistralai/voxtral-small-24b-2507":{"id":"mistralai/voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32768},"cost":{"input":0.1058,"output":0.3174,"cache_read":0.01058},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Small 3.2 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07935,"output":0.2116},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":102400},"cost":{"input":2.116,"output":6.348,"cache_read":0.2116},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Small 3.1 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.371358,"output":0.58719},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.3174,"output":0.3174,"cache_read":0.01587},"canonical_model_id":"mistral/ministral-8b-2512"},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":26214},"cost":{"input":0.2116,"output":0.6348,"cache_read":0.02116},"canonical_model_id":"mistral/mistral-saba"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.0529,"output":0.08464},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.198375,"output":0.7935,"cache_read":0.01587},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral NeMo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1587,"output":0.17986,"cache_read":0.1587},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-medium-3-5":{"id":"mistralai/mistral-medium-3-5","name":"Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.587,"output":7.935},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4232,"output":2.116,"cache_read":0.04232},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1587,"output":0.1587,"cache_read":0.01058},"canonical_model_id":"mistral/ministral-3b-2512"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6348,"output":1.9044,"cache_read":0.11638},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"cost":{"input":0.13225,"output":0.529,"cache_read":0.01058},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM 4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3174,"output":0.9522,"cache_read":0.05819},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":117964},"cost":{"input":0.1587,"output":0.48668,"cache_read":0.07935},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B Instruct","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":1.04742,"output":1.57642,"cache_read":1.04742},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":96000},"cost":{"input":0.0529,"output":0.1587,"cache_read":0.04232},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.5},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048756,"output":65536},"cost":{"input":2.116,"output":12.696,"cache_read":0.2116,"cache_write":0.39675},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.2645,"output":1.587,"cache_read":0.02645,"cache_write":0.088167},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":32768},"cost":{"input":0.0529,"output":0.1058,"cache_read":0.04232},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-8-flash":{"id":"google/gemini-3-8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.587,"output":9.522,"cache_read":0.1587,"cache_write":0.088167},"canonical_model_id":"google/gemini-flash-latest"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.1587,"output":0.6348,"cache_read":0.1587},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemma-2-27b-it":{"id":"google/gemma-2-27b-it","name":"Gemma 2 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-07-13","last_updated":"2024-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0.6877,"output":0.6877},"canonical_model_id":"google/gemma-2-27b-it"},"inception/mercury-2":{"id":"inception/mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.2645,"output":0.7935,"cache_read":0.02645},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.2116,"output":0.7935,"cache_read":0.02116},"canonical_model_id":"inception/mercury-2.5"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.14812,"output":0.29624,"cache_read":0.002962},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.46023,"output":0.92046,"cache_read":0.003809},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1.058,"output":2.8566,"cache_read":0.0529},"canonical_model_id":"stepfun/step-5-preview"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"cost":{"input":0.61364,"output":1.52352,"cache_read":0.122728},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.058,"output":4.2849,"cache_read":0.17986},"canonical_model_id":"thinkingmachines/inkling"},"mancer/weaver":{"id":"mancer/weaver","name":"Weaver (alpha)","description":"General-purpose chat model for instruction following, writing, and analysis","family":"alpha","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-08-02","last_updated":"2023-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":2000},"cost":{"input":0.7935,"output":1.058},"canonical_model_id":"mancer/weaver"},"unbiased/pareto":{"id":"unbiased/pareto","name":"Pareto","description":"Blended multimodal model that runs multiple language models in parallel and returns a synthesized answer","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.645,"output":7.935,"cache_read":0.2645},"canonical_model_id":"unbiased/pareto"},"relace/relace-apply-3":{"id":"relace/relace-apply-3","name":"Apply 3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.8993,"output":1.3225},"canonical_model_id":"relace/relace-apply-3"},"relace/relace-search":{"id":"relace/relace-search","name":"Search","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":1.058,"output":3.174},"canonical_model_id":"relace/relace-search"},"tencent/hy-mt2-7b":{"id":"tencent/hy-mt2-7b","name":"Hy-MT2-7B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.078292,"output":0.31211},"canonical_model_id":"tencent/hy-mt2-7b"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.882372,"output":2.646058,"cache_read":0.044436},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy-mt2-30b-a3b":{"id":"tencent/hy-mt2-30b-a3b","name":"Hy-MT2-30B-A3B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.078292,"output":0.31211},"canonical_model_id":"tencent/hy-mt2-30b-a3b"},"tencent/hy-mt2-1.8b":{"id":"tencent/hy-mt2-1.8b","name":"Hy-MT2-1.8B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.046552,"output":0.187266},"canonical_model_id":"tencent/hy-mt2-1.8b"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.2116,"output":0.8464,"cache_read":0.0529},"canonical_model_id":"tencent/hy3"},"meta-llama/Llama-3.3-70B-Instruct-Turbo":{"id":"meta-llama/Llama-3.3-70B-Instruct-Turbo","name":"Llama 3.3 70B Instruct Turbo","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":127000},"cost":{"input":1.04,"output":1.04},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.9,"output":2.7,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.84092,"output":3.68184,"cache_read":0.15341},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-chat-v3-0324":{"id":"deepseek/deepseek-chat-v3-0324","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":124000},"cost":{"input":0.30682,"output":1.20612,"cache_read":0.14812},"canonical_model_id":"deepseek/deepseek-chat-v3-0324"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.39,"output":1.56,"cache_read":0.0078},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"nousresearch/hermes-3-llama-3.1-70b":{"id":"nousresearch/hermes-3-llama-3.1-70b","name":"Hermes 3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.7406,"output":0.7406},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"nousresearch/hermes-3-llama-3.1-405b":{"id":"nousresearch/hermes-3-llama-3.1-405b","name":"Hermes 3 405B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1.058,"output":1.058},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"rekaai/reka-edge":{"id":"rekaai/reka-edge","name":"Reka Edge","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.1058,"output":0.1058},"canonical_model_id":"rekaai/reka-edge"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":1.058,"output":3.174},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6348,"output":2.645,"cache_read":0.1587},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.60306,"output":2.4334},"canonical_model_id":"moonshotai/kimi-k2"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3703,"output":1.587,"cache_read":0.04232},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.3225,"output":4.4965,"cache_read":0.1587},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.3225,"output":4.4965,"cache_read":0.1587},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.3225,"output":4.4965,"cache_read":0.1587},"canonical_model_id":"meta/muse-spark-1.2"},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":100000},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"alibaba/qwen-plus":{"id":"alibaba/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131000,"output":16384},"cost":{"input":0.4,"output":1.2,"cache_read":0.08},"canonical_model_id":"alibaba/qwen-plus"},"alibaba/qwen3.8-max-prime":{"id":"alibaba/qwen3.8-max-prime","name":"Qwen3.8 Max Prime","description":"High-throughput edition of Qwen3.8 Max for coding, professional work, multimodal understanding, and long-running agent workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4.232,"output":12.696,"cache_read":0.529},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"alibaba/qwen-max":{"id":"alibaba/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":1.6,"output":6.4,"cache_read":0.32},"canonical_model_id":"alibaba/qwen-max"},"alibaba/qwen3.8-2.4t-a95b":{"id":"alibaba/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":52429},"cost":{"input":2.645,"output":6.6125,"cache_read":0.529},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"alibaba/qwen3.8-flash":{"id":"alibaba/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"alibaba/qwen3-next-80b-a3b-instruct":{"id":"alibaba/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":129024,"output":16384},"cost":{"input":0.1587,"output":1.2696},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"alibaba/qwen3-coder-480b-a35b-instruct":{"id":"alibaba/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":1.058,"output":5.15775,"cache_read":1.058},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"alibaba/qwen3.8-27b":{"id":"alibaba/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.4761,"output":3.3856},"canonical_model_id":"alibaba/qwen3.8-27b"},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.1587,"output":0.6348,"cache_read":0.01587},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.03174,"output":0.12696,"cache_read":0.006348},"canonical_model_id":"upstage/solar-pro4"},"cognitivecomputations/dolphin-mistral-24b-venice-edition":{"id":"cognitivecomputations/dolphin-mistral-24b-venice-edition","name":"Venice Uncensored","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2116,"output":0.9522},"canonical_model_id":"cognitivecomputations/dolphin-mistral-24b-venice-edition"},"ibm-granite/granite-4.0-h-micro":{"id":"ibm-granite/granite-4.0-h-micro","name":"Granite 4.0 Micro","description":"Compact open-weight hybrid Granite model for lightweight enterprise chat and tool calling","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.017986,"output":0.118496},"canonical_model_id":"ibm/granite-4-h-micro"},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"Aion-RP 1.0 (8B)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.8464,"output":1.6928},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"amazon/nova-micro-v1":{"id":"amazon/nova-micro-v1","name":"Nova Micro 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":5120},"cost":{"input":0.03703,"output":0.14812},"canonical_model_id":"amazon/nova-micro-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.8464,"output":3.3856},"canonical_model_id":"amazon/nova-pro-v1"},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.06348,"output":0.25392},"canonical_model_id":"amazon/nova-lite-v1"},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"output":4096},"cost":{"input":0.529,"output":0.7935},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"openai/gpt-5-2":{"id":"openai/gpt-5-2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5-5":{"id":"openai/gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":117964},"cost":{"input":0.3703,"output":1.0051,"cache_read":0.3703},"canonical_model_id":"openai/gpt-oss-120b"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":158.7,"output":634.8},"canonical_model_id":"openai/o1-pro"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1-Codex-Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2.645,"output":21.16,"cache_read":0.2645},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna Pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2116,"output":1.2696,"cache_read":0.02116},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":117964},"cost":{"input":0.07935,"output":0.3174,"cache_read":0.039675},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5-4":{"id":"openai/gpt-5-4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":22.218,"output":177.744},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.587,"output":9.522,"cache_read":0.1587},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":31.74,"output":190.44},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":3.703,"output":29.624,"cache_read":0.3703},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3 Mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra Pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.116,"output":12.696,"cache_read":0.2116},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":31.74,"output":190.44},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.4-image-2":{"id":"openai/gpt-5.4-image-2","type":"image","name":"GPT-5.4 Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":8.464,"output":15.87,"cache_read":2.116},"canonical_model_id":"openai/gpt-5.4-image-2"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15.87,"output":126.96},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"gpt-oss-safeguard-20b","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.07935,"output":0.3174,"cache_read":0.039675},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2116,"output":1.3225,"cache_read":0.02116},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":8192},"cost":{"input":30,"output":60,"cache_read":30},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.232,"output":21.16,"cache_read":0.4232},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o Mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3 Pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":21.16,"output":84.64},"canonical_model_id":"openai/o3-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5-1":{"id":"openai/gpt-5-1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03174,"output":0.1587},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.0529,"output":0.24334},"canonical_model_id":"inference-net/schematron-v2-small"},"thedrummer/skyfall-36b-v2":{"id":"thedrummer/skyfall-36b-v2","name":"Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.5819,"output":0.8464,"cache_read":0.2645},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"thedrummer/cydonia-24b-v4.1":{"id":"thedrummer/cydonia-24b-v4.1","name":"Cydonia 24B V4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3174,"output":0.529,"cache_read":0.1587},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"thedrummer/unslopnemo-12b":{"id":"thedrummer/unslopnemo-12b","name":"UnslopNemo 12B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-11-08","last_updated":"2024-11-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.4232,"output":0.4232},"canonical_model_id":"thedrummer/unslopnemo-12b"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude 4.7 Opus","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":26.45,"cache_read":0.529,"cache_write":6.6125},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude 4.5 Haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.058,"output":5.29,"cache_read":0.1058,"cache_write":1.3225},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude 4.5 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.29,"output":26.45,"cache_read":0.529,"cache_write":6.6125},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude 4.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.174,"output":15.87,"cache_read":0.3174,"cache_write":3.9675},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude 4.8 Opus","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":26.45,"cache_read":0.529,"cache_write":6.6125},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude 4.6 Sonnet","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.174,"output":15.87,"cache_read":0.3174,"cache_write":3.9675},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax 13B","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":1.98375,"output":1.98375,"cache_read":1.98375},"canonical_model_id":"gryphe/mythomax-l2-13b"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":31.74,"cache_read":0.529},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.116,"output":6.348,"cache_read":0.2645},"canonical_model_id":"sakana/fugu-max"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.29,"output":31.74,"cache_read":0.529},"canonical_model_id":"sakana/fugu-ultra"},"sao10k/l3-lunaris-8b":{"id":"sao10k/l3-lunaris-8b","name":"Llama 3 8B Lunaris","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-13","last_updated":"2024-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":7372},"cost":{"input":0.0529,"output":0.0529},"canonical_model_id":"sao10k/l3-lunaris-8b"},"sao10k/l3.1-euryale-70b":{"id":"sao10k/l3.1-euryale-70b","name":"Llama 3.1 Euryale 70B v2.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1.534523,"output":1.534523},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"sao10k/l3.3-euryale-70b":{"id":"sao10k/l3.3-euryale-70b","name":"Llama 3.3 Euryale 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-12-18","last_updated":"2024-12-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.6877,"output":0.7935},"canonical_model_id":"sao10k/l3.3-euryale-70b"},"x-ai/grok-4-fast-non-reasoning":{"id":"x-ai/grok-4-fast-non-reasoning","name":"Grok 4 Fast Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1999000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"x-ai/grok-4-20-0309-non-reasoning":{"id":"x-ai/grok-4-20-0309-non-reasoning","name":"Grok 4.20 Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1999000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"Seed-2.0-Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.2645,"output":2.116},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-mini":{"id":"bytedance-seed/seed-2.0-mini","name":"Seed-2.0-Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1058,"output":0.4232},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance-seed/seed-1.6-flash":{"id":"bytedance-seed/seed-1.6-flash","name":"Seed 1.6 Flash","description":"Low-latency ByteDance Seed model for high-throughput chat, extraction, and lightweight tool use","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.07935,"output":0.3174},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"Seed-2.0-Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.529,"output":3.174},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":16384},"cost":{"input":3.174,"output":5.29},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"zhipu/glm-5.2":{"id":"zhipu/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zhipu/glm-5.3":{"id":"zhipu/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B (12-2024)","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.039675,"output":0.1587},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+ (08-2024)","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.645,"output":10.58},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R (08-2024)","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.1587,"output":0.6348},"canonical_model_id":"cohere/command-r-08-2024"},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":2.645,"output":10.58},"canonical_model_id":"cohere/command-a"},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"V3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.9522,"output":2.0102},"canonical_model_id":"morph/morph-v3-large"},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"V3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":81920,"output":38000},"cost":{"input":0.8464,"output":1.2696},"canonical_model_id":"morph/morph-v3-fast"}}},"daoxe":{"id":"daoxe","env":["DAOXE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://daoxe.com/v1","name":"DaoXE","doc":"https://daoxe.com/pricing","models":{"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":4,"output":16,"cache_read":0.8},"canonical_model_id":"moonshotai/kimi-k2.5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":5},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"}}},"kimi-code-plan-global":{"id":"kimi-code-plan-global","env":["KIMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.kimi.ai/coding/v1","name":"Kimi For Coding (kimi.ai)","doc":"https://www.kimi.ai/code/docs/en/kimi-code/models.html","models":{"k3-256k":{"id":"k3-256k","name":"Kimi K3-256K","description":"256K-context version of Kimi K3, reducing token consumption for shorter coding sessions","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"k3.1-pro":{"id":"k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"k3":{"id":"k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-for-coding-highspeed":{"id":"kimi-for-coding-highspeed","name":"Kimi For Coding HighSpeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"k3.1-pro-highspeed":{"id":"k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-for-coding":{"id":"kimi-for-coding","name":"kimi-for-coding","description":"Kimi coding model with more efficient thinking and up to 1M context, available through Kimi Code","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.8-preview"}}},"boundless":{"id":"boundless","env":["BOUNDLESS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inference.boundless.network/v1","name":"Boundless","doc":"https://inference.boundless.network/models","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.34,"output":2.3,"cache_read":0.17},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":1,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.056},"canonical_model_id":"minimax/MiniMax-M3"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-5"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01},"canonical_model_id":"openai/gpt-6-luna"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"openai/gpt-6-sol"},"claude-fable-5.1":{"id":"claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25},"canonical_model_id":"anthropic/claude-fable-5-1"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.3,"output":11.4,"cache_read":0.23},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.04,"output":2.08,"cache_read":0.0352},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"claude-opus-5.5":{"id":"claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"anthropic/claude-opus-5-5"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"dsv4":{"id":"dsv4","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.04,"output":0.21,"cache_read":0.02},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.6":{"id":"qwen3.6","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.04,"output":0.35,"cache_read":0.02},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.12,"output":3.52,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"nemotron3-super":{"id":"nemotron3-super","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.12,"output":0.52,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.12,"output":0.4,"cache_read":0.024},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-sol"}}},"volcengine":{"id":"volcengine","env":["ARK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ark.cn-beijing.volces.com/api/v3","name":"Volcengine Ark","doc":"https://www.volcengine.com/docs/82379/1330310","models":{"doubao-seed-2-1-pro-260628":{"id":"doubao-seed-2-1-pro-260628","name":"Seed 2.1 Pro","description":"Flagship ByteDance Seed 2.1 model for complex multimodal reasoning, coding, and agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.8906,"output":4.45301,"cache_read":0.17812},"canonical_model_id":"bytedance-seed/seed-2.1-pro"},"doubao-seed-2-0-pro-260215":{"id":"doubao-seed-2-0-pro-260215","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.47499,"output":2.37494,"cache_read":0.095,"tiers":[{"input":0.71248,"output":3.56241,"cache_read":0.1425,"tier":{"type":"context","size":32000}},{"input":1.42496,"output":7.12482,"cache_read":0.28499,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"doubao-seed-2-1-turbo-260628":{"id":"doubao-seed-2-1-turbo-260628","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.4453,"output":2.22651,"cache_read":0.08906},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"doubao-seed-2-0-lite-260428":{"id":"doubao-seed-2-0-lite-260428","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.08906,"output":0.53436,"cache_read":0.01781,"tiers":[{"input":0.13359,"output":0.80154,"cache_read":0.02672,"tier":{"type":"context","size":32000}},{"input":0.26718,"output":1.60308,"cache_read":0.05344,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"deepseek-v4-flash-ga-260731":{"id":"deepseek-v4-flash-ga-260731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.4453,"output":1.3359,"cache_read":0.01484},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"doubao-seed-1-6-251015":{"id":"doubao-seed-1-6-251015","name":"Seed 1.6","description":"ByteDance Seed model for long-context reasoning, instruction following, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":2.37494,"cache_read":0.02375,"tier":{"type":"context","size":32000}},{"input":0.35624,"output":3.56241,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"doubao-seed-character-260628":{"id":"doubao-seed-character-260628","name":"Seed Character","description":"ByteDance Seed model optimized for character-driven dialogue and consistent conversational behavior","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.11875,"output":0.29687,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":0.8906,"cache_read":0.02375,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"bytedance-seed/seed-character"},"doubao-seed-evolving":{"id":"doubao-seed-evolving","name":"Seed Evolving","description":"Rolling ByteDance Seed model for rapidly updated reasoning, coding, and agent capabilities","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.8906,"output":4.45301,"cache_read":0.17812},"canonical_model_id":"bytedance-seed/seed-evolving"},"doubao-seed-2-0-code-preview-260215":{"id":"doubao-seed-2-0-code-preview-260215","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.47499,"output":2.37494,"cache_read":0.095,"tiers":[{"input":0.71248,"output":3.56241,"cache_read":0.1425,"tier":{"type":"context","size":32000}},{"input":1.42496,"output":7.12482,"cache_read":0.28499,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"doubao-seed-1-6-flash-250828":{"id":"doubao-seed-1-6-flash-250828","name":"Seed 1.6 Flash","description":"Low-latency ByteDance Seed model for high-throughput chat, extraction, and lightweight tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.02227,"output":0.22265,"cache_read":0.00445,"tiers":[{"input":0.04453,"output":0.4453,"cache_read":0.00445,"tier":{"type":"context","size":32000}},{"input":0.08906,"output":0.8906,"cache_read":0.00445,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"glm-5-2-260617":{"id":"glm-5-2-260617","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.18747,"output":4.15615,"cache_read":0.29687},"canonical_model_id":"zhipuai/glm-5.2"},"doubao-seed-1-6-vision-250815":{"id":"doubao-seed-1-6-vision-250815","name":"Seed 1.6 Vision","description":"ByteDance Seed multimodal model for image understanding, visual reasoning, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-08-15","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":2.37494,"cache_read":0.02375,"tier":{"type":"context","size":32000}},{"input":0.35624,"output":3.56241,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-vision"},"deepseek-v4-pro-ga-260813":{"id":"deepseek-v4-pro-ga-260813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.3359,"output":4.00771,"cache_read":0.04453},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"doubao-seed-2-0-mini-260428":{"id":"doubao-seed-2-0-mini-260428","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.02969,"output":0.29687,"cache_read":0.00594,"tiers":[{"input":0.05937,"output":0.59374,"cache_read":0.01187,"tier":{"type":"context","size":32000}},{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"glm-5-3-flash-260828":{"id":"glm-5-3-flash-260828","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11875,"output":0.41563,"cache_read":0.03414},"canonical_model_id":"zhipuai/glm-5.3-flash"},"doubao-seed-1-8-251228":{"id":"doubao-seed-1-8-251228","name":"Seed 1.8","description":"ByteDance Seed model for multimodal reasoning, long-context analysis, and agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.11875,"output":1.18747,"cache_read":0.02375,"tiers":[{"input":0.17812,"output":2.37494,"cache_read":0.02375,"tier":{"type":"context","size":32000}},{"input":0.35624,"output":3.56241,"cache_read":0.02375,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-8"}}},"commandcode":{"id":"commandcode","env":["CMD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.commandcode.ai/provider/v1","name":"CommandCode","doc":"https://commandcode.ai/docs/provider","models":{"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":0},"canonical_model_id":"openai/gpt-5.4"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":0},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0},"canonical_model_id":"openai/gpt-5.4-mini"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2,"output":8,"cache_read":0.5,"cache_write":0},"canonical_model_id":"openai/gpt-5.3-codex"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"openai/gpt-5.6-sol"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"cache_write":0.08334},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"stepfun/Step-3.7-Flash":{"id":"stepfun/Step-3.7-Flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/Step-5-Preview":{"id":"stepfun/Step-5-Preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"stepfun/Step-3.5-Flash":{"id":"stepfun/Step-3.5-Flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.09,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":0.5,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy3-paid":{"id":"tencent/hy3-paid","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash-fast":{"id":"deepseek/deepseek-v4-flash-fast","name":"DeepSeek V4 Flash Fast","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-fast":{"id":"deepseek/deepseek-v4.1-flash-fast","name":"DeepSeek V4.1 Flash Fast","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.16,"output":0.58,"cache_read":0.016},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.7-Code-Highspeed":{"id":"moonshotai/Kimi-K2.7-Code-Highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 (Contributor)","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 (Contributor)","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":3,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.8-Omni-Flash":{"id":"Qwen/Qwen3.8-Omni-Flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"Qwen/Qwen3.8-Max-0902":{"id":"Qwen/Qwen3.8-Max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"Qwen/Qwen3.8-Max":{"id":"Qwen/Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"Qwen/Qwen3.8-Flash":{"id":"Qwen/Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"inclusionai/ling-3.1-flash:free":{"id":"inclusionai/ling-3.1-flash:free","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.1-flash"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"mistral/mistral-large-4":{"id":"mistral/mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"zai-org/GLM-5.2-Fast":{"id":"zai-org/GLM-5.2-Fast","name":"GLM-5.2-Fast","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":10.25,"cache_read":0.5},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"inco":{"id":"inco","env":["INCO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inco.ai/v1","name":"Inco","doc":"https://platform.inco.ai/docs","models":{"kimi-k3:fast":{"id":"kimi-k3:fast","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":6,"output":30,"cache_read":0.6},"canonical_model_id":"moonshotai/kimi-k3"},"glm-5.3-flash:fast":{"id":"glm-5.3-flash:fast","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1,"cache_read":0.06},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5.3:fast":{"id":"glm-5.3:fast","name":"GLM-5.3 Fast","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.52},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-v4.1-flash:fast":{"id":"deepseek-v4.1-flash:fast","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.6,"output":2.4,"cache_read":0.012},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax-m3:fast":{"id":"minimax-m3:fast","name":"MiniMax M3 Fast","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"}}},"berget":{"id":"berget","env":["BERGET_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.berget.ai/v1","name":"Berget.AI","doc":"https://api.berget.ai","models":{"mistralai/Mistral-Small-3.2-24B-Instruct-2506":{"id":"mistralai/Mistral-Small-3.2-24B-Instruct-2506","name":"Mistral Small 3.2 24B Instruct 2506","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":8192},"cost":{"input":0.33645,"output":0.33645},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B Instruct","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["audio","image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.280375,"output":0.56075},"canonical_model_id":"google/gemma-4-31b-it"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"output":32768},"cost":{"input":3.3645,"output":16.8225,"cache_read":1.00935},"canonical_model_id":"moonshotai/kimi-k3"},"Qwen/Qwen3.8-27B-FP8":{"id":"Qwen/Qwen3.8-27B-FP8","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.46,"output":3.48},"canonical_model_id":"alibaba/qwen3.8-27b"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"cost":{"input":0.29,"output":0.58},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":32768},"status":"deprecated","cost":{"input":1.54,"output":4.84},"canonical_model_id":"zhipuai/glm-5.2"}}},"kosmik":{"id":"kosmik","env":["KOSMIK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.koscompute.com/v1","name":"Kosmik Compute","doc":"https://api.koscompute.com/docs/","models":{"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.35,"output":2.2,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"moonshotai-cn":{"id":"moonshotai-cn","env":["MOONSHOT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.moonshot.cn/v1","name":"Moonshot AI (China)","doc":"https://platform.moonshot.cn/docs/api/chat","models":{"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.9711,"output":4.0337,"cache_read":0.1942},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.9437,"output":10.1671,"cache_read":0.1944,"cache_write":1.9437},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":2.9879,"output":14.9395,"cache_read":0.2988,"cache_write":2.9879},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.9711,"output":4.0337,"cache_read":0.1643},"canonical_model_id":"moonshotai/kimi-k2.6"},"kimi-k3.1-pro-highspeed":{"id":"kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":3.8874,"output":20.3342,"cache_read":0.1944,"cache_write":3.8874},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code HighSpeed","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9421,"output":8.0673,"cache_read":0.3884},"canonical_model_id":"moonshotai/kimi-k2.7-code"}}},"kimi-code-plan-cn":{"id":"kimi-code-plan-cn","env":["KIMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.kimi.com/coding/v1","name":"Kimi For Coding (kimi.com)","doc":"https://www.kimi.com/code/docs/en/kimi-code/models.html","models":{"k3-256k":{"id":"k3-256k","name":"Kimi K3-256K","description":"256K-context version of Kimi K3, reducing token consumption for shorter coding sessions","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"k3.1-pro":{"id":"k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"k3":{"id":"k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-for-coding-highspeed":{"id":"kimi-for-coding-highspeed","name":"Kimi For Coding HighSpeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"k3.1-pro-highspeed":{"id":"k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-for-coding":{"id":"kimi-for-coding","name":"Kimi K2.8 Preview","description":"Kimi coding model with more efficient thinking and up to 1M context, available through Kimi Code","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.8-preview"}}},"venice":{"id":"venice","env":["VENICE_API_KEY"],"npm":"venice-ai-sdk-provider","name":"Venice AI","doc":"https://docs.venice.ai","models":{"openai-gpt-6-luna":{"id":"openai-gpt-6-luna","name":"GPT-6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"z-ai-glm-5-3-flash":{"id":"z-ai-glm-5-3-flash","name":"GLM 5.3 Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-3-6-flash":{"id":"gemini-3-6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.9375,"output":4.6875,"cache_read":0.09375},"canonical_model_id":"google/gemini-3.6-flash"},"qwen3-5-35b-a3b":{"id":"qwen3-5-35b-a3b","name":"Qwen 3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.15,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen-3-8-max":{"id":"qwen-3-8-max","name":"Qwen 3.8 Max","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.3125,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"openai-gpt-6-astra":{"id":"openai-gpt-6-astra","name":"GPT-6 Astra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2.5,"output":12.5,"cache_read":0.125},"canonical_model_id":"google/gemini-4-argon"},"mercury-2":{"id":"mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.3125,"output":0.9375,"cache_read":0.03125},"canonical_model_id":"inception/mercury-2"},"aion-labs-aion-3-5-mini":{"id":"aion-labs-aion-3-5-mini","name":"Aion 3.5 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.875,"output":1.75,"cache_read":0.225}},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.75,"output":3.5,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":0.7,"output":3.75,"cache_read":0.07},"canonical_model_id":"google/gemini-3-flash-preview"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":198000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"abliteration-abliterated-model-large-flash":{"id":"abliteration-abliterated-model-large-flash","name":"Abliterated Large Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.125}},"openai-gpt-56-sol-pro":{"id":"openai-gpt-56-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"mistral-small-3-2-24b-instruct":{"id":"mistral-small-3-2-24b-instruct","name":"Mistral Small 3.2 24B Instruct","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.09375,"output":0.25},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"openai-gpt-6-sol":{"id":"openai-gpt-6-sol","name":"GPT-6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"minimax-m3-1-flash":{"id":"minimax-m3-1-flash","name":"MiniMax M3.1 Flash","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1875,"output":0.75,"cache_read":0.01875},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"claude-opus-4-8-fast":{"id":"claude-opus-4-8-fast","name":"Claude Opus 4.8 Fast","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"aion-labs-aion-3-5":{"id":"aion-labs-aion-3-5","name":"Aion 3.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":3.75,"output":7.5,"cache_read":0.9375}},"qwen3-next-80b":{"id":"qwen3-next-80b","name":"Qwen 3 Next 80b","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.35,"output":1.9},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"gemini-3-1-pro-preview":{"id":"gemini-3-1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2.5,"output":15,"cache_read":0.5,"cache_write":0.5,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"cache_write":0.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5,"cache_write":0.5}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"qwen-3-7-max":{"id":"qwen-3-7-max","name":"Qwen 3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2.7,"output":8.05,"cache_read":0.27,"cache_write":3.35},"canonical_model_id":"alibaba/qwen3.7-max"},"grok-4-7":{"id":"grok-4-7","name":"Grok 4.7","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":200000},"cost":{"input":2.27,"output":6.8,"cache_read":0.57,"tiers":[{"input":4.53,"output":13.6,"cache_read":1.13,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.53,"output":13.6,"cache_read":1.13}},"canonical_model_id":"xai/grok-4.7"},"llama-3.3-70b":{"id":"llama-3.3-70b","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.7,"output":2.8},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":1.65,"output":3.301,"cache_read":0.33},"canonical_model_id":"deepseek/deepseek-v4-pro"},"kimi-k3-1-pro":{"id":"kimi-k3-1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.30000000000000004},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"nvidia-nemotron-3-nano-30b-a3b":{"id":"nvidia-nemotron-3-nano-30b-a3b","name":"NVIDIA Nemotron 3 Nano 30B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.075,"output":0.3},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"zai-org-glm-4.7":{"id":"zai-org-glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":16384},"cost":{"input":0.55,"output":2.65,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai-glm-5v-turbo":{"id":"z-ai-glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32768},"cost":{"input":1.5,"output":5,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5v-turbo"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"qwen-3-7-plus":{"id":"qwen-3-7-plus","name":"Qwen 3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":2,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":1.5,"output":6,"cache_read":0.15,"cache_write":1.875,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.5,"output":6,"cache_read":0.15,"cache_write":1.875}},"canonical_model_id":"alibaba/qwen3.7-plus"},"kimi-k2-5":{"id":"kimi-k2-5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.56,"output":3.5,"cache_read":0.22},"canonical_model_id":"moonshotai/kimi-k2.5"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 Preview","description":"Preview model for early access evaluation, prototyping, and compatibility testing","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":1.0425,"output":3.12625,"cache_read":0.0525},"canonical_model_id":"tencent/hy4-preview"},"grok-4-3":{"id":"grok-4-3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":1.42,"output":2.83,"cache_read":0.23,"tiers":[{"input":2.83,"output":5.67,"cache_read":0.45,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.83,"output":5.67,"cache_read":0.45}},"canonical_model_id":"xai/grok-4.3"},"gemini-3-7-flash":{"id":"gemini-3-7-flash","name":"Gemini 3.7 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.9375,"output":4.6875,"cache_read":0.09375},"canonical_model_id":"google/gemini-3.7-flash"},"abliteration-abliterated-model-large-v2":{"id":"abliteration-abliterated-model-large-v2","name":"Abliterated Large V2","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":3,"output":5,"cache_read":0.3}},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"qwen3-coder-480b-a35b-instruct-turbo":{"id":"qwen3-coder-480b-a35b-instruct-turbo","name":"Qwen 3 Coder 480B Turbo","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-01-27","last_updated":"2026-06-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.35,"output":1.5,"cache_read":0.04}},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"openai-gpt-56-luna":{"id":"openai-gpt-56-luna","name":"GPT-5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai-gpt-54":{"id":"openai-gpt-54","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":131072},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai-gpt-56-terra-pro":{"id":"openai-gpt-56-terra-pro","name":"GPT-5.6 Terra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"qwen-3-8-2-4t-a95b":{"id":"qwen-3-8-2-4t-a95b","name":"Qwen 3.8 2.4T","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":2.5,"output":7.5,"cache_read":0.3125},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"aion-labs-aion-3-0-mini":{"id":"aion-labs-aion-3-0-mini","name":"Aion 3.0 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":0.875,"output":1.75,"cache_read":0.225},"canonical_model_id":"aion-labs/aion-3.0-mini"},"kimi-k2-7-code":{"id":"kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.75,"output":3.5,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.0075},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"grok-4-5":{"id":"grok-4-5","name":"Grok 4.5","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":32000},"cost":{"input":2.27,"output":6.8,"cache_read":0.34,"tiers":[{"input":4.53,"output":13.6,"cache_read":0.68,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.53,"output":13.6,"cache_read":0.68}},"canonical_model_id":"xai/grok-4.5"},"zai-org-glm-4.6":{"id":"zai-org-glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":16384},"cost":{"input":0.43,"output":1.75,"cache_read":0.08},"canonical_model_id":"zhipuai/glm-4.6"},"openai-gpt-6-astra-pro":{"id":"openai-gpt-6-astra-pro","name":"GPT-6 Astra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai-gpt-56-luna-pro":{"id":"openai-gpt-56-luna-pro","name":"GPT-5.6 Luna Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen-3-8-27b":{"id":"qwen-3-8-27b","name":"Qwen 3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.45,"output":3.2},"canonical_model_id":"alibaba/qwen3.8-27b"},"olafangensan-glm-4.7-flash-heretic":{"id":"olafangensan-glm-4.7-flash-heretic","name":"GLM 4.7 Flash Heretic","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-02-04","last_updated":"2026-06-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":24000},"cost":{"input":0.07,"output":0.4,"cache_read":0.035}},"openai-gpt-61-sol":{"id":"openai-gpt-61-sol","name":"GPT-6.1 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"xiaomi-mimo-v2-5":{"id":"xiaomi-mimo-v2-5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2,"cache_read":0.08},"canonical_model_id":"xiaomi/mimo-v2.5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"openai-gpt-52":{"id":"openai-gpt-52","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":272000,"output":65536},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"qwen3-6-35b-a3b":{"id":"qwen3-6-35b-a3b","name":"Qwen 3.6 35B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.1,"output":1},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gemini-3-5-flash":{"id":"gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.55,"output":9.45,"cache_read":0.155,"cache_write":0.086},"canonical_model_id":"google/gemini-3.5-flash"},"openai-gpt-61-sol-ultrafast":{"id":"openai-gpt-61-sol-ultrafast","name":"GPT-6.1 Sol Ultrafast","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":12,"output":60,"cache_read":0.6,"cache_write":15,"tiers":[{"input":24,"output":90,"cache_read":1.2,"cache_write":30,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":24,"output":90,"cache_read":1.2,"cache_write":30}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-5-fast":{"id":"claude-opus-5-fast","name":"Claude Opus 5 Fast","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-5"},"aion-labs-aion-3-0":{"id":"aion-labs-aion-3-0","name":"Aion 3.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":3.75,"output":7.5,"cache_read":0.9375},"canonical_model_id":"aion-labs/aion-3.0"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"openai-gpt-55":{"id":"openai-gpt-55","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"deepseek-v4-1-flash-1017":{"id":"deepseek-v4-1-flash-1017","name":"DeepSeek v4.1 Flash (1017)","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.0075},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"grok-build-0-1":{"id":"grok-build-0-1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3.75,"output":18.75,"cache_read":0.375},"canonical_model_id":"moonshotai/kimi-k3"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen 3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.75},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":1.65,"output":4.95,"cache_read":0.165},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"google-gemma-4-31b-it":{"id":"google-gemma-4-31b-it","name":"Google Gemma 4 31B Instruct","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.12,"output":0.36,"cache_read":0.09},"canonical_model_id":"google/gemma-4-31b-it"},"openai-gpt-56-terra":{"id":"openai-gpt-56-terra","name":"GPT-5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gemini-3-5-flash-lite":{"id":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.375,"output":3.125,"cache_read":0.0375},"canonical_model_id":"google/gemini-3.5-flash-lite"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"zai-org-glm-5":{"id":"zai-org-glm-5","name":"GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":32000},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"venice-uncensored-role-play":{"id":"venice-uncensored-role-play","name":"Venice Role Play Uncensored","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"venice","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-02-20","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.5,"output":2}},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"qwen3-5-397b-a17b":{"id":"qwen3-5-397b-a17b","name":"Qwen 3.5 397B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.75,"output":4.5},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-v4-1-flash-uncensored":{"id":"deepseek-v4-1-flash-uncensored","name":"DeepSeek V4.1 Flash Uncensored","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.33,"output":1.32,"cache_read":0.00825}},"zai-org-glm-5-2":{"id":"zai-org-glm-5-2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":200000},"cost":{"input":2.27,"output":6.8,"cache_read":0.57,"tiers":[{"input":4.53,"output":13.6,"cache_read":1.13,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.53,"output":13.6,"cache_read":1.13}},"canonical_model_id":"xai/grok-4.6"},"kimi-k3-fast-api":{"id":"kimi-k3-fast-api","name":"Kimi K3 Fast","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"grok-4-20":{"id":"grok-4-20","name":"Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-03-12","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":1.42,"output":2.83,"cache_read":0.23,"tiers":[{"input":2.83,"output":5.67,"cache_read":0.45,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.83,"output":5.67,"cache_read":0.45}}},"minimax-m25":{"id":"minimax-m25","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":32768},"cost":{"input":0.27,"output":0.95,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"openai-gpt-oss-120b":{"id":"openai-gpt-oss-120b","name":"OpenAI GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-120b"},"openai-gpt-4o-mini-2024-07-18":{"id":"openai-gpt-4o-mini-2024-07-18","name":"GPT-4o Mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"qwen-3-6-plus":{"id":"qwen-3-6-plus","name":"Qwen 3.6 Plus Uncensored","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.625,"output":3.75,"cache_read":0.0625,"cache_write":0.78,"tiers":[{"input":2.5,"output":7.5,"cache_read":0.0625,"cache_write":0.78,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2.5,"output":7.5,"cache_read":0.0625,"cache_write":0.78}},"canonical_model_id":"alibaba/qwen3.6-plus"},"muse-wonder":{"id":"muse-wonder","name":"Muse Wonder","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-12","last_updated":"2026-10-15","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":5,"output":25,"cache_read":0.25}},"qwen-3-8-flash":{"id":"qwen-3-8-flash","name":"Qwen 3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.14,"output":0.49,"cache_read":0.014},"canonical_model_id":"alibaba/qwen3.8-flash"},"z-ai-glm-5-3":{"id":"z-ai-glm-5-3","name":"GLM 5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.75,"output":5.5,"cache_read":0.325},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3-235b-a22b-thinking-2507":{"id":"qwen3-235b-a22b-thinking-2507","name":"Qwen 3 235B A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.45,"output":3.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"xiaomi-mimo-v2-6-flash":{"id":"xiaomi-mimo-v2-6-flash","name":"MiMo-V2.6-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.175,"output":0.35,"cache_read":0.00375},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"kimi-k3-1-pro-highspeed":{"id":"kimi-k3-1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":6,"output":30,"cache_read":0.6000000000000001}},"minimax-m3-preview":{"id":"minimax-m3-preview","name":"MiniMax M3 Preview","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax-m3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"release_date":"2026-06-12","last_updated":"2026-06-13","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.3,"output":1.2,"cache_read":0.06}},"gemini-3-8-flash":{"id":"gemini-3-8-flash","name":"Gemini 3.8 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.9375,"output":4.6875,"cache_read":0.09375},"canonical_model_id":"google/gemini-3.8-flash"},"openai-gpt-54-pro":{"id":"openai-gpt-54-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"llama-3.2-3b":{"id":"llama-3.2-3b","name":"Llama 3.2 3B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"meta/llama-3.2-3b"},"google-gemma-3-27b-it":{"id":"google-gemma-3-27b-it","name":"Google Gemma 3 27B Instruct","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":16384},"cost":{"input":0.12,"output":0.2},"canonical_model_id":"google/gemma-3-27b-it"},"nvidia-nemotron-3-ultra-550b-a55b":{"id":"nvidia-nemotron-3-ultra-550b-a55b","name":"NVIDIA Nemotron 3 Ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0.625,"output":3.125,"cache_read":0.1875},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":198000,"output":32768},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"openai-gpt-55-pro":{"id":"openai-gpt-55-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"deepseek-v4-flash-0731-fast":{"id":"deepseek-v4-flash-0731-fast","name":"DeepSeek V4 Flash 0731 Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.35,"output":0.7,"cache_read":0.0875},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"gemma-4-uncensored":{"id":"gemma-4-uncensored","name":"Gemma 4 Uncensored","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-04-13","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.1625,"output":0.5}},"zai-org-glm-5-1":{"id":"zai-org-glm-5-1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":80000},"cost":{"input":1.54,"output":4.84,"cache_read":0.286},"canonical_model_id":"zhipuai/glm-5.1"},"claude-opus-5-5-fast":{"id":"claude-opus-5-5-fast","name":"Claude Opus 5.5 Fast","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":8,"output":40,"cache_read":0.4,"cache_write":10},"canonical_model_id":"anthropic/claude-opus-5-5"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":160000,"output":32768},"status":"deprecated","cost":{"input":0.33,"output":0.48,"cache_read":0.16},"canonical_model_id":"deepseek/deepseek-v3.2"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"seed-2-1-turbo":{"id":"seed-2-1-turbo","name":"Seed 2.1 Turbo","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":0.625,"output":3.125,"cache_read":0.125},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"openai-gpt-54-mini":{"id":"openai-gpt-54-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash 0423","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.138,"output":0.275,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"mercury-3":{"id":"mercury-3","name":"Mercury 3","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01},"canonical_model_id":"inception/mercury-3"},"grok-4-20-multi-agent":{"id":"grok-4-20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"release_date":"2026-03-12","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":1.42,"output":2.83,"cache_read":0.23,"tiers":[{"input":2.83,"output":5.67,"cache_read":0.45,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.83,"output":5.67,"cache_read":0.45}}},"hermes-3-llama-3.1-405b":{"id":"hermes-3-llama-3.1-405b","name":"Hermes 3 Llama 3.1 405b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":1.1,"output":3},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"google-gemma-4-26b-a4b-it":{"id":"google-gemma-4-26b-a4b-it","name":"Google Gemma 4 26B A4B Instruct","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.13,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"status":"deprecated","cost":{"input":0.5625,"output":1.5,"cache_read":0.125},"canonical_model_id":"thinkingmachines/inkling"},"z-ai-glm-5-turbo":{"id":"z-ai-glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32768},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"nvidia-nemotron-3-5-super-vl-120b-a12b":{"id":"nvidia-nemotron-3-5-super-vl-120b-a12b","name":"NVIDIA Nemotron 3.5 Super VL 120B","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.25,"output":1.25,"cache_read":0.125},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"openai-gpt-56-sol":{"id":"openai-gpt-56-sol","name":"GPT-5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai-gpt-4o-2024-11-20":{"id":"openai-gpt-4o-2024-11-20","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"mercury-2-5":{"id":"mercury-2-5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.25,"output":0.9375,"cache_read":0.025},"canonical_model_id":"inception/mercury-2.5"},"minimax-m27":{"id":"minimax-m27","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":32768},"cost":{"input":0.375,"output":1.5,"cache_read":0.06875},"canonical_model_id":"minimax/MiniMax-M2.7"},"venice-uncensored-1-2":{"id":"venice-uncensored-1-2","name":"Venice Uncensored 1.2","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"venice","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-04-01","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2,"output":0.9}},"openai-gpt-53-codex":{"id":"openai-gpt-53-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"Qwen3 VL 235B","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-01-16","last_updated":"2026-06-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.21,"output":1.9,"cache_read":0.1}},"qwen3-6-27b":{"id":"qwen3-6-27b","name":"Qwen 3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.325,"output":3.25},"canonical_model_id":"alibaba/qwen3.6-27b"},"muse-spark-open":{"id":"muse-spark-open","name":"Muse Spark Open","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.5625,"output":5.3125,"cache_read":0.1875},"canonical_model_id":"meta/muse-spark-open"}}},"infer":{"id":"infer","env":["INFER_API_KEY"],"npm":"@ai-sdk/openai","api":"https://infer.flow7.org/v1","name":"Infer by Flow7","doc":"https://infer.flow7.org/opencode","models":{"infer/gpt-5.6-sol:official":{"id":"infer/gpt-5.6-sol:official","name":"GPT-5.6 Sol (Official API)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":271999,"output":128000},"provider":{"shape":"responses"},"cost":{"input":2.5,"output":12.5,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-sol"},"infer/gpt-6-astra:official":{"id":"infer/gpt-6-astra:official","name":"GPT-6 Astra (Official API)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":271999,"output":128000},"provider":{"shape":"responses"},"cost":{"input":12.5,"output":62.5,"cache_read":1.25,"cache_write":15.625},"canonical_model_id":"openai/gpt-6-astra"}}},"qvac":{"id":"qvac","env":["QVAC_API_KEY"],"npm":"@qvac/ai-sdk-provider","name":"QVAC","doc":"https://www.npmjs.com/package/@qvac/ai-sdk-provider","models":{"qwen3.5-0.8b":{"id":"qwen3.5-0.8b","name":"Qwen3.5 0.8B","description":"Qwen instruction model for multilingual chat and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-0.8b"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-9b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-120b"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-20b"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"gemma4-31b":{"id":"gemma4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"qwen3.5-4b":{"id":"qwen3.5-4b","name":"Qwen3.5 4B","description":"Qwen instruction model for multilingual chat and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-4b"},"qwen3.5-2b":{"id":"qwen3.5-2b","name":"Qwen3.5 2B","description":"Qwen instruction model for multilingual chat and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-2b"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"inferx":{"id":"inferx","env":["INFERX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://model.inferx.net/endpoints/v1","name":"InferX","doc":"https://model.inferx.net/endpoints","models":{"Devstral-2-123B-Instruct-2512-int4-AutoRound":{"id":"Devstral-2-123B-Instruct-2512-int4-AutoRound","name":"Devstral-2-123B-Instruct-2512-int4-AutoRound","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"mistral/devstral-2512"},"Qwen3.6-35B-A3B-FP8":{"id":"Qwen3.6-35B-A3B-FP8","name":"Qwen3.6 35B A3B FP8","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Agents-A1":{"id":"Agents-A1","name":"Agents-A1","description":"35B MoE agentic model built for long-horizon search, engineering, and scientific reasoning tasks","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"release_date":"2026-06-26","last_updated":"2026-06-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":100000},"cost":{"input":0,"output":0}},"Ornith-1.0-35B-FP8":{"id":"Ornith-1.0-35B-FP8","name":"Ornith-1.0-35B-FP8","description":"Large coding-reasoning model for agentic software tasks and RL search","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-25","last_updated":"2026-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":100000},"cost":{"input":0,"output":0},"canonical_model_id":"deepreinforce/ornith-1.0-35b"},"Qwen3.6-27B-FP8":{"id":"Qwen3.6-27B-FP8","name":"Qwen3.6 27B FP8","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"gemma-4-31B-it-fp8":{"id":"gemma-4-31B-it-fp8","name":"Gemma 4 31B IT FP8","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"Qwen3-Coder-Next-FP8":{"id":"Qwen3-Coder-Next-FP8","name":"Qwen3 Coder Next FP8","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-next"},"Qwen3-Coder-Next-FP8-no-thinking":{"id":"Qwen3-Coder-Next-FP8-no-thinking","name":"Qwen3-Coder-Next-FP8-no-thinking","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":260000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-next"},"Qwen3.6-35B-A3B-fp8-no-thinking":{"id":"Qwen3.6-35B-A3B-fp8-no-thinking","name":"Qwen3.6-35B-A3B-fp8-no-thinking","description":"Qwen3.6-35B-A3B-fp8 disable thinking","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"deepseek-v4-flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":100000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"Qwen3-Embedding-8B":{"id":"Qwen3-Embedding-8B","type":"embedding","name":"Qwen3-Embedding-8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"mimo-v25":{"id":"mimo-v25","name":"mimo-v25","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":100000},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5"}}},"google":{"id":"google","env":["GOOGLE_API_KEY","GOOGLE_GENERATIVE_AI_API_KEY","GEMINI_API_KEY"],"npm":"@ai-sdk/google","name":"Google","doc":"https://ai.google.dev/gemini-api/docs/models","models":{"deep-research-max-preview-04-2026":{"id":"deep-research-max-preview-04-2026","name":"Deep Research Max Preview (Apr-21-2026)","description":"Maximum-comprehensiveness agentic researcher for multi-step investigation, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/deep-research-max-preview-04-2026"},"lyria-3-pro-preview":{"id":"lyria-3-pro-preview","name":"Lyria 3 Pro Preview","description":"Music generation model for full-length songs from text or images with vocals and structure","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-pro-preview"},"gemini-embedding-001":{"id":"gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"cost":{"input":0.15,"output":0}},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":60},"canonical_model_id":"google/gemini-3.1-flash-image"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768}},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"gemini-2.5-computer-use-preview-10-2025":{"id":"gemini-2.5-computer-use-preview-10-2025","name":"Gemini 2.5 Computer Use Preview 10-2025","description":"Specialized Gemini 2.5 model for browser-control agents that automate UI tasks","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-10-07","last_updated":"2025-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":1.25,"output":10,"tiers":[{"input":2.5,"output":15,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15}},"canonical_model_id":"google/gemini-2.5-computer-use-preview-10-2025"},"gemini-2.5-flash-preview-tts":{"id":"gemini-2.5-flash-preview-tts","type":"speech","name":"Gemini 2.5 Flash Preview TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-01","last_updated":"2025-05-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"cost":{"input":0.5,"output":10}},"gemini-flash-lite-latest":{"id":"gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"gemini-3.1-flash-tts-preview":{"id":"gemini-3.1-flash-tts-preview","type":"speech","name":"Gemini 3.1 Flash TTS Preview","description":"Low-latency speech generation with steerable prompts and expressive audio tags","family":"gemini-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-15","last_updated":"2026-04-15","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"cost":{"input":1,"output":20},"canonical_model_id":"google/gemini-3.1-flash-tts-preview"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1}},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5}},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"gemini-2.5-pro-preview-tts":{"id":"gemini-2.5-pro-preview-tts","type":"speech","name":"Gemini 2.5 Pro Preview TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-01","last_updated":"2025-05-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"cost":{"input":1,"output":20}},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}}},"gemini-3.1-pro-preview-customtools":{"id":"gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}}},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"lyria-3-clip-preview":{"id":"lyria-3-clip-preview","name":"Lyria 3 Clip Preview","description":"Music generation model for short 30-second clips, loops, and previews from text or image prompts","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-clip-preview"},"deep-research-preview-04-2026":{"id":"deep-research-preview-04-2026","name":"Deep Research Preview (Apr-21-2026)","description":"Agentic model for autonomous multi-step research, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/deep-research-preview-04-2026"},"gemini-3.1-flash-live-preview":{"id":"gemini-3.1-flash-live-preview","type":"realtime","name":"Gemini 3.1 Flash Live Preview","description":"High-quality, low-latency Live API model for real-time dialogue and voice-first AI applications","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-26","last_updated":"2026-03-26","modalities":{"input":["text","image","video","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":0.75,"output":4.5,"input_audio":3,"output_audio":12},"canonical_model_id":"google/gemini-3.1-flash-live-preview"},"gemini-3.1-flash-lite-image":{"id":"gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":30},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30,"cache_read":0.075},"canonical_model_id":"google/gemini-2.5-flash-image"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Legacy model retained for compatibility with older integrations","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5}},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"gemini-pro-latest":{"id":"gemini-pro-latest","name":"Gemini Pro Latest","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini-omni-flash-preview":{"id":"gemini-omni-flash-preview","type":"video","name":"Gemini Omni Flash Preview","description":"Video generation and editing model for fast, conversational text- and image-to-video workflows","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":1.5,"output":17.5},"canonical_model_id":"google/gemini-omni-flash-preview"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"gemini-3.5-live-translate-preview":{"id":"gemini-3.5-live-translate-preview","type":"realtime","name":"Gemini 3.5 Live Translate Preview","description":"Low-latency audio-to-audio model for real-time speech translation across 70+ languages","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["audio"],"output":["audio","text"]},"open_weights":false,"limit":{"context":16384,"output":32768},"cost":{"input":3.5,"output":21,"input_audio":3.5,"output_audio":21},"canonical_model_id":"google/gemini-3.5-live-translate-preview"},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image-preview"},"gemini-flash-latest":{"id":"gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75}},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768}},"gemini-embedding-2":{"id":"gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1},"cost":{"input":0.2,"output":0,"input_audio":6.5},"canonical_model_id":"google/gemini-embedding-2"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5}},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3}},"gemini-3.1-flash-image-preview":{"id":"gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":60}}}},"nvidia":{"id":"nvidia","env":["NVIDIA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://integrate.api.nvidia.com/v1","name":"Nvidia","doc":"https://docs.api.nvidia.com/nim/","models":{"mistralai/magistral-small-2506":{"id":"mistralai/magistral-small-2506","name":"Magistral Small 2506","description":"Open Mistral reasoning model for transparent step-by-step problem solving","family":"magistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/magistral-small-2506"},"mistralai/mistral-medium-3-instruct":{"id":"mistralai/mistral-medium-3-instruct","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/mistral-7b-instruct-v0.3":{"id":"mistralai/mistral-7b-instruct-v0.3","name":"Mistral-7B-Instruct-v0.3","description":"Open 7B Mistral instruction model with function calling","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-22","last_updated":"2024-05-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-7b-instruct-v0.3"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma-4-31B-IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-3-12b-it"},"google/diffusiongemma-26b-a4b-it":{"id":"google/diffusiongemma-26b-a4b-it","name":"DiffusionGemma 26B A4B IT","description":"Gemma 4 based discrete text diffusion model for fast chat, reasoning, and tool use","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":250000,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/diffusiongemma-26b-a4b-it"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-3-4b-it"},"google/google-paligemma":{"id":"google/google-paligemma","name":"paligemma","description":"Vision-language model for answering questions about images and text","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-05-14","last_updated":"2024-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"nvidia/cosmos-reason2-8b":{"id":"nvidia/cosmos-reason2-8b","name":"Cosmos Reason2 8B","description":"Vision language model for physical-world understanding with structured reasoning on video and images","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/nemotron-3.5-super-vl-120b-a12b":{"id":"nvidia/nemotron-3.5-super-vl-120b-a12b","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3.5-lightning-30b-a3b":{"id":"nvidia/nemotron-3.5-lightning-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/studiovoice":{"id":"nvidia/studiovoice","name":"studiovoice","description":"Enhances noisy or reverberant speech recordings to studio-quality speech","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-03","last_updated":"2025-06-13","modalities":{"input":["audio"],"output":["audio"]},"open_weights":true,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0}},"nvidia/streampetr":{"id":"nvidia/streampetr","name":"streampetr","description":"Temporal 3D object detection model for autonomous driving","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"nvidia/llama-3.1-nemotron-70b-instruct":{"id":"nvidia/llama-3.1-nemotron-70b-instruct","name":"Llama 3.1 Nemotron 70B Instruct","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-3.1-nemotron-70b-instruct"},"nvidia/usdvalidate":{"id":"nvidia/usdvalidate","name":"usdvalidate","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-07-24","last_updated":"2025-01-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":4096},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/llama-3.1-nemotron-safety-guard-8b-v3":{"id":"nvidia/llama-3.1-nemotron-safety-guard-8b-v3","name":"llama-3.1-nemotron-safety-guard-8b-v3","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-3.1-nemotron-safety-guard-8b-v3"},"nvidia/usdcode":{"id":"nvidia/usdcode","name":"usdcode","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-01","last_updated":"2026-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-voicechat":{"id":"nvidia/nemotron-voicechat","name":"nemotron-voicechat","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-voicechat"},"nvidia/active-speaker-detection":{"id":"nvidia/active-speaker-detection","name":"Active Speaker Detection","description":"Detects and tracks which person is speaking across video frames","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/synthetic-video-detector":{"id":"nvidia/synthetic-video-detector","name":"synthetic-video-detector","description":"Detects AI-generated (synthetic) video","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/llama-nemotron-embed-vl-1b-v2":{"id":"nvidia/llama-nemotron-embed-vl-1b-v2","type":"embedding","name":"llama-nemotron-embed-vl-1b-v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"nemotron","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-nemotron-embed-vl-1b-v2"},"nvidia/magpie-tts-zeroshot":{"id":"nvidia/magpie-tts-zeroshot","type":"speech","name":"magpie-tts-zeroshot","description":"Expressive text-to-speech that clones a voice from a short audio sample","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-22","last_updated":"2025-06-12","modalities":{"input":["text","audio"],"output":["audio"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/riva-translate-4b-instruct-v1.1":{"id":"nvidia/riva-translate-4b-instruct-v1.1","name":"riva-translate-4b-instruct-v1_1","description":"Translation model for multilingual conversion, localization, and cross-language workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-12","last_updated":"2025-12-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0,"output":0}},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]},{"type":"budget_tokens","min":-1,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/rerank-qa-mistral-4b":{"id":"nvidia/rerank-qa-mistral-4b","type":"reranking","name":"rerank-qa-mistral-4b","description":"Reranking model for improving retrieval quality in search and recommendation systems","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-17","last_updated":"2025-01-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"api":"https://ai.api.nvidia.com/v1/retrieval/nvidia/reranking"},"cost":{"input":0,"output":0}},"nvidia/llama-3.1-nemotron-ultra-253b-v1":{"id":"nvidia/llama-3.1-nemotron-ultra-253b-v1","name":"Llama 3.1 Nemotron Ultra 253B","description":"Flagship Nemotron model for high-throughput reasoning and complex agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-04-07","last_updated":"2025-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-3.1-nemotron-ultra-253b"},"nvidia/llama-nemotron-rerank-vl-1b-v2":{"id":"nvidia/llama-nemotron-rerank-vl-1b-v2","type":"reranking","name":"llama-nemotron-rerank-vl-1b-v2","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"nemotron","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10240,"output":4096},"provider":{"api":"https://ai.api.nvidia.com/v1/retrieval/nvidia/llama-nemotron-rerank-vl-1b-v2/reranking"},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/llama-nemotron-rerank-vl-1b-v2"},"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning":{"id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning","name":"Nemotron 3 Nano Omni","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":-1,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/cosmos-transfer2_5-2b":{"id":"nvidia/cosmos-transfer2_5-2b","type":"video","name":"cosmos-transfer2.5-2b","description":"Generates physics-aware video world states from text and spatial control videos","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":true,"limit":{"context":0,"output":4096},"cost":{"input":0,"output":0}},"nvidia/bevformer":{"id":"nvidia/bevformer","name":"bevformer","description":"Multi-frame bird's-eye-view 3D perception transformer for autonomous driving","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-03-18","last_updated":"2025-07-20","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"nvidia/sparsedrive":{"id":"nvidia/sparsedrive","name":"sparsedrive","description":"End-to-end autonomous driving stack for perception, prediction, and planning","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-03-18","last_updated":"2025-07-20","modalities":{"input":["video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/llama-3.2-90b-vision-instruct":{"id":"meta/llama-3.2-90b-vision-instruct","name":"Llama-3.2-90B-Vision-Instruct","description":"Large open multimodal Llama model for image reasoning, document QA, and captioning","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.2-90b-vision-instruct"},"meta/llama-3.2-11b-vision-instruct":{"id":"meta/llama-3.2-11b-vision-instruct","name":"Llama 3.2 11b Vision Instruct","description":"Open multimodal Llama model for image understanding, captioning, and visual QA","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"meta/llama-guard-4-12b":{"id":"meta/llama-guard-4-12b","name":"Llama Guard 4 12B","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-guard-4-12b"},"black-forest-labs/flux_1-kontext-dev":{"id":"black-forest-labs/flux_1-kontext-dev","type":"image","name":"FLUX.1-Kontext-dev","description":"Open-weight FLUX.1 Kontext model for instruction-based image editing","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-29","last_updated":"2025-05-29","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.1-kontext-dev"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-1-kontext-dev"},"black-forest-labs/flux_2-klein-4b":{"id":"black-forest-labs/flux_2-klein-4b","type":"image","name":"FLUX.2 Klein 4B","description":"Fast 4B open-weight image generation and editing model","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-06","release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.2-klein-4b"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-2-klein-4b"},"black-forest-labs/flux_1-schnell":{"id":"black-forest-labs/flux_1-schnell","type":"image","name":"FLUX.1-schnell","description":"Fast few-step FLUX.1 text-to-image model under Apache 2.0","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-07","release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.1-schnell"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-1-schnell"},"black-forest-labs/flux.1-dev":{"id":"black-forest-labs/flux.1-dev","type":"image","name":"FLUX.1-dev","description":"Open-weight FLUX.1 text-to-image model for high-quality guidance-distilled generation","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-08","release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.1-dev"},"cost":{"input":0,"output":0},"canonical_model_id":"black-forest-labs/flux-1-dev"},"deepseek-ai/deepseek-v4.1-flash":{"id":"deepseek-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-20b"},"openai/whisper-large-v3":{"id":"openai/whisper-large-v3","type":"transcription","name":"Whisper Large v3","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-09","release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"openai/whisper-large-v3"},"stepfun-ai/step-5-preview":{"id":"stepfun-ai/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":1048576},"cost":{"input":0,"output":0},"canonical_model_id":"stepfun/step-5-preview"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"qwen/qwen-image-edit":{"id":"qwen/qwen-image-edit","type":"image","name":"Qwen Image Edit","description":"Image editing model with multilingual text editing and subject consistency","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-08-19","last_updated":"2025-08-19","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0}},"qwen/qwen-image":{"id":"qwen/qwen-image","type":"image","name":"Qwen Image","description":"Text-to-image foundation model with multilingual text rendering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0}},"minimaxai/minimax-m3.1-flash":{"id":"minimaxai/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"}}},"alibaba-cn":{"id":"alibaba-cn","env":["DASHSCOPE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://dashscope.aliyuncs.com/compatible-mode/v1","name":"Alibaba (China)","doc":"https://www.alibabacloud.com/help/en/model-studio/models","models":{"qwen-turbo":{"id":"qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":16384},"status":"deprecated","cost":{"input":0.044,"output":0.087,"reasoning":0.431}},"deepseek-v3-1":{"id":"deepseek-v3-1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0.574,"output":1.721}},"qwen-long":{"id":"qwen-long","name":"Qwen Long","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-25","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"output":8192},"cost":{"input":0.072,"output":0.287}},"deepseek-r1-distill-qwen-1-5b":{"id":"deepseek-r1-distill-qwen-1-5b","name":"DeepSeek R1 Distill Qwen 1.5B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"cost":{"input":0,"output":0}},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.1,"output":3.851,"cache_read":0.275,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.276,"output":1.101,"cache_read":0.0552,"cache_write":0.345,"tiers":[{"input":0.826,"output":3.301,"cache_read":0.1652,"cache_write":1.0325,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.826,"output":3.301,"cache_read":0.1652,"cache_write":1.0325}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qvq-max":{"id":"qvq-max","name":"QVQ Max","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qvq","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":1.147,"output":4.588}},"qwen-plus":{"id":"qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.115,"output":0.287,"reasoning":1.147,"cache_read":0.023,"cache_write":0.144,"tiers":[{"input":0.345,"output":2.868,"reasoning":3.441,"cache_read":0.069,"cache_write":0.43125,"tier":{"type":"context","size":128000}},{"input":0.689,"output":6.881,"reasoning":9.175,"cache_read":0.1378,"cache_write":0.86125,"tier":{"type":"context","size":256000}}]}},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.216,"output":0.861,"tiers":[{"input":0.323,"output":1.291,"tier":{"type":"context","size":32000}},{"input":0.538,"output":2.151,"tier":{"type":"context","size":128000}}]}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.29754,"output":1.19015,"cache_read":0.01488},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"moonshot-kimi-k2-instruct":{"id":"moonshot-kimi-k2-instruct","name":"Moonshot Kimi K2 Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"deepseek-r1-distill-qwen-7b":{"id":"deepseek-r1-distill-qwen-7b","name":"DeepSeek R1 Distill Qwen 7B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.072,"output":0.144}},"qwen-math-turbo":{"id":"qwen-math-turbo","name":"Qwen Math Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":3072},"status":"deprecated","cost":{"input":0.287,"output":0.861}},"qwen-mt-plus":{"id":"qwen-mt-plus","name":"Qwen-MT Plus","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":0.259,"output":0.775}},"qwen-max":{"id":"qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.345,"output":1.377}},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.65,"output":3.301},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen-mt-turbo":{"id":"qwen-mt-turbo","name":"Qwen-MT Turbo","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"status":"deprecated","cost":{"input":0.101,"output":0.28}},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-14","last_updated":"2026-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0.825,"output":3.301,"cache_read":0.165,"cache_write":1.03125,"tiers":[{"input":1.1,"output":3.851,"cache_read":0.22,"cache_write":1.375,"tier":{"type":"context","size":32000}}]}},"deepseek-r1-distill-qwen-32b":{"id":"deepseek-r1-distill-qwen-32b","name":"DeepSeek R1 Distill Qwen 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.287,"output":0.861}},"qwen-omni-turbo":{"id":"qwen-omni-turbo","name":"Qwen-Omni Turbo","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-19","last_updated":"2025-03-26","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.058,"output":0.23,"input_audio":3.584,"output_audio":7.168}},"qwen-plus-character":{"id":"qwen-plus-character","name":"Qwen Plus Character","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"cost":{"input":0.115,"output":0.287}},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.143353,"output":1.433525,"cache_read":0.028671,"cache_write":0.179191,"tiers":[{"input":0.215,"output":2.15,"cache_read":0.043,"cache_write":0.26875,"tier":{"type":"context","size":32000}},{"input":0.43,"output":4.301,"cache_read":0.086,"cache_write":0.5375,"tier":{"type":"context","size":128000}}]}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Moonshot Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.574,"output":3.011,"cache_read":0.1148,"cache_write":0.7175}},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11875,"output":0.40073,"cache_read":0.014844,"cache_write":0.185547},"canonical_model_id":"alibaba/qwen3.8-flash"},"deepseek-r1-distill-llama-8b":{"id":"deepseek-r1-distill-llama-8b","name":"DeepSeek R1 Distill Llama 8B","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0,"output":0}},"qwen3.8-omni-flash":{"id":"qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.113,"output":0.382,"cache_read":0.014},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.276,"output":1.651,"cache_read":0.0276,"cache_write":0.345,"tiers":[{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625}}},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.144,"output":0.574}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":2.827,"output":14.133,"cache_read":0.283},"canonical_model_id":"moonshotai/kimi-k3"},"qwen-deep-research":{"id":"qwen-deep-research","name":"Qwen Deep Research","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":7.742,"output":23.367}},"qwen-doc-turbo":{"id":"qwen-doc-turbo","name":"Qwen Doc Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.087,"output":0.144}},"kimi-k2.6":{"id":"kimi-k2.6","name":"Moonshot Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.929,"output":3.858,"cache_read":0.1858,"cache_write":1.16125}},"qwen-vl-ocr":{"id":"qwen-vl-ocr","name":"Qwen-VL OCR","description":"OCR model for extracting structured text from documents and screenshots","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-10-28","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":34096,"output":4096},"cost":{"input":0.043,"output":0.072}},"deepseek-v3-2-exp":{"id":"deepseek-v3-2-exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0.287,"output":0.431}},"qwen3-omni-flash-realtime":{"id":"qwen3-omni-flash-realtime","type":"realtime","name":"Qwen3-Omni Flash Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"status":"deprecated","provider":{"api":"wss://dashscope.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.23,"output":0.918,"input_audio":3.584,"output_audio":7.168}},"qwen3.5-flash":{"id":"qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-09-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.029,"output":0.287,"reasoning":0.287,"tiers":[{"input":0.115,"output":1.147,"reasoning":1.147,"tier":{"type":"context","size":128000}},{"input":0.172,"output":1.72,"reasoning":1.72,"tier":{"type":"context","size":256000}}]}},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-09-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.115,"output":0.688,"reasoning":0.688,"tiers":[{"input":0.287,"output":1.72,"reasoning":1.72,"tier":{"type":"context","size":128000}},{"input":0.573,"output":3.44,"reasoning":3.44,"tier":{"type":"context","size":256000}}]}},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.172,"output":1.032,"reasoning":1.032,"tiers":[{"input":0.43,"output":2.58,"reasoning":2.58,"tier":{"type":"context","size":128000}}]}},"qwen3-14b":{"id":"qwen3-14b","name":"Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.144,"output":0.574,"reasoning":1.434}},"qwen3-asr-flash":{"id":"qwen3-asr-flash","type":"transcription","name":"Qwen3-ASR Flash","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-04","release_date":"2025-09-08","last_updated":"2025-09-08","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":53248,"output":4096}},"qwen-vl-max":{"id":"qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.23,"output":0.574}},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.165,"output":0.99,"cache_read":0.0165,"cache_write":0.20625,"tiers":[{"input":0.66,"output":3.961,"cache_read":0.066,"cache_write":0.825,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.66,"output":3.961,"cache_read":0.066,"cache_write":0.825}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.861,"output":3.441,"tiers":[{"input":1.291,"output":5.161,"tier":{"type":"context","size":32000}},{"input":2.151,"output":8.602,"tier":{"type":"context","size":128000}}]}},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-23","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.359,"output":1.434,"reasoning":1.434,"tiers":[{"input":0.574,"output":2.294,"reasoning":2.294,"tier":{"type":"context","size":32000}},{"input":1.004,"output":4.014,"reasoning":4.014,"tier":{"type":"context","size":128000}}]}},"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-11","last_updated":"2026-02-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0.573,"output":2.58,"tiers":[{"input":0.86,"output":3.154,"tier":{"type":"context","size":32000}}]}},"qwen3-vl-30b-a3b":{"id":"qwen3-vl-30b-a3b","name":"Qwen3-VL 30B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.108,"output":0.431,"reasoning":1.076}},"qwen-omni-turbo-realtime":{"id":"qwen-omni-turbo-realtime","type":"realtime","name":"Qwen-Omni Turbo Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-05-08","last_updated":"2025-05-08","modalities":{"input":["text","image","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","provider":{"api":"wss://dashscope.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.23,"output":0.918,"input_audio":3.584,"output_audio":7.168}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2026-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.574,"output":2.296,"tiers":[{"input":0.861,"output":3.444,"tier":{"type":"context","size":32000}},{"input":1.435,"output":5.74,"tier":{"type":"context","size":128000}},{"input":2.87,"output":28.7,"tier":{"type":"context","size":256000}}]}},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1,"output":3.851,"cache_read":0.275,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"qwq-plus":{"id":"qwq-plus","name":"QwQ Plus","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.23,"output":0.574}},"deepseek-r1-0528":{"id":"deepseek-r1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.287,"output":1.147,"reasoning":2.868}},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.65,"output":4.951,"cache_read":0.33,"cache_write":2.0625},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen3-next-80b-a3b-thinking":{"id":"qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.144,"output":1.434}},"deepseek-r1-distill-llama-70b":{"id":"deepseek-r1-distill-llama-70b","name":"DeepSeek R1 Distill Llama 70B","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"cost":{"input":0.287,"output":0.861}},"qwen-flash":{"id":"qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.022,"output":0.216,"cache_read":0.0044,"cache_write":0.0275,"tiers":[{"input":0.087,"output":0.861,"cache_read":0.0174,"cache_write":0.10875,"tier":{"type":"context","size":128000}},{"input":0.173,"output":1.721,"cache_read":0.0346,"cache_write":0.21625,"tier":{"type":"context","size":256000}}]}},"qwen2-5-omni-7b":{"id":"qwen2-5-omni-7b","name":"Qwen2.5-Omni 7B","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-12","last_updated":"2024-12","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":true,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.087,"output":0.345,"input_audio":5.448}},"qwen-vl-plus":{"id":"qwen-vl-plus","name":"Qwen-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.115,"output":0.287}},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen3 235B-A22B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.287,"output":1.147,"reasoning":2.868}},"tongyi-intent-detect-v3":{"id":"tongyi-intent-detect-v3","name":"Tongyi Intent Detect V3","description":"General-purpose chat model for instruction following, writing, and analysis","family":"yi","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1024},"cost":{"input":0.058,"output":0.144}},"deepseek-v3":{"id":"deepseek-v3","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-12-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"status":"deprecated","cost":{"input":0.287,"output":1.147}},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Moonshot Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"qwen3-omni-flash":{"id":"qwen3-omni-flash","name":"Qwen3-Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"cost":{"input":0.258,"output":0.989,"input_audio":2.265,"output_audio":8.974}},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.77744,"output":5.33231,"cache_read":0.22218,"cache_write":2.22179},"canonical_model_id":"alibaba/qwen3.8-max"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"qwen-math-plus":{"id":"qwen-math-plus","name":"Qwen Math Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-08-16","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":3072},"status":"deprecated","cost":{"input":0.574,"output":1.721}},"qwen3.6-max-preview":{"id":"qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":245800,"output":65536},"status":"deprecated","cost":{"input":1.32,"output":7.9,"cache_read":0.132}},"qwen3-8b":{"id":"qwen3-8b","name":"Qwen3 8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":38912}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.072,"output":0.287,"reasoning":0.717}},"qwen3-coder-flash":{"id":"qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.144,"output":0.574,"cache_read":0.0288,"cache_write":0.18,"tiers":[{"input":0.216,"output":0.861,"cache_read":0.0432,"cache_write":0.27,"tier":{"type":"context","size":32000}},{"input":0.359,"output":1.434,"cache_read":0.0718,"cache_write":0.44875,"tier":{"type":"context","size":128000}},{"input":0.717,"output":3.584,"cache_read":0.1434,"cache_write":0.89625,"tier":{"type":"context","size":256000}}]}},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":131072},"cost":{"input":0.02962,"output":0.1185,"cache_read":0.005924,"cache_write":0.03703,"tiers":[{"input":0.08887,"output":0.35549,"cache_read":0.017774,"cache_write":0.11109,"tier":{"type":"context","size":32000}},{"input":0.17774,"output":0.71098,"cache_read":0.035548,"cache_write":0.22218,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"Qwen3-VL 235B-A22B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.286705,"output":1.14682,"reasoning":2.867051}},"deepseek-r1-distill-qwen-14b":{"id":"deepseek-r1-distill-qwen-14b","name":"DeepSeek R1 Distill Qwen 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.144,"output":0.431}},"deepseek-r1":{"id":"deepseek-r1","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.574,"output":2.294}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2}},"MiniMax/MiniMax-M2.7":{"id":"MiniMax/MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375}},"kimi/kimi-k2.5":{"id":"kimi/kimi-k2.5","name":"kimi/kimi-k2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1}},"siliconflow/deepseek-v3.1-terminus":{"id":"siliconflow/deepseek-v3.1-terminus","name":"siliconflow/deepseek-v3.1-terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":65536},"cost":{"input":0.27,"output":1}},"siliconflow/deepseek-v3-0324":{"id":"siliconflow/deepseek-v3-0324","name":"siliconflow/deepseek-v3-0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":163840},"cost":{"input":0.25,"output":1}},"siliconflow/deepseek-r1-0528":{"id":"siliconflow/deepseek-r1-0528","name":"siliconflow/deepseek-r1-0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.18}},"siliconflow/deepseek-v3.2":{"id":"siliconflow/deepseek-v3.2","name":"siliconflow/deepseek-v3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-03","last_updated":"2025-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":65536},"cost":{"input":0.27,"output":0.42}}}},"inception":{"id":"inception","env":["INCEPTION_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inceptionlabs.ai/v1/","name":"Inception","doc":"https://docs.inceptionlabs.ai/get-started/models","models":{"mercury-2":{"id":"mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"mercury-2.5":{"id":"mercury-2.5","name":"Mercury 2.5","description":"Mercury 2.5 is the fastest reasoning LLM, and the latest diffusion LLM (dLLM) from Inception","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-11-01","release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.2,"output":0.75,"cache_read":0.02},"canonical_model_id":"inception/mercury-2.5"},"mercury-edit-2":{"id":"mercury-edit-2","name":"Mercury Edit 2","description":"Code editing dLLM for autocomplete (FIM) and next-edit suggestions","family":"mercury","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-30","last_updated":"2026-03-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-edit-2"},"mercury-3":{"id":"mercury-3","name":"Mercury 3","description":"Inception's diffusion reasoning model for chat, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01},"canonical_model_id":"inception/mercury-3"}}},"xiaomi":{"id":"xiaomi","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.xiaomimimo.com/v1","name":"Xiaomi","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"MiMo Pro model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.6-pro-ultraspeed":{"id":"mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5-pro-ultraspeed":{"id":"mimo-v2.5-pro-ultraspeed","name":"MiMo-V2.5-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-06-08","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.305,"output":2.61,"cache_read":0.0108},"canonical_model_id":"xiaomi/mimo-v2.5-pro-ultraspeed"}}},"coralbricks":{"id":"coralbricks","env":["CORAL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.coralbricks.ai/v1","name":"CoralBricks","doc":"https://www.coralbricks.ai/docs","models":{"deepseek-v4.1-flash-fast":{"id":"deepseek-v4.1-flash-fast","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.25,"output":1.2,"cache_read":0.005,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"}}},"model-oracle-ai":{"id":"model-oracle-ai","env":["MODEL_ORACLE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.modeloracle.com/api/v1","name":"Model Oracle AI","doc":"https://modeloracle.com/setup/","models":{"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1-mini"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-5.4"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-pro"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-5.5"},"claude-haiku-4.5":{"id":"claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"canonical_model_id":"anthropic/claude-fable-5"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.4-mini"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o4-mini"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.4-nano"},"auto":{"id":"auto","name":"Auto","description":"Model Oracle AI decision engine that selects and routes among configured coding-agent models","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-29","last_updated":"2026-07-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5"},"claude-opus-4.8":{"id":"claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"canonical_model_id":"anthropic/claude-opus-4-8"}}},"stepfun":{"id":"stepfun","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.com/v1","name":"StepFun (China)","doc":"https://platform.stepfun.com/docs/zh/overview/concept","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.20175,"output":1.21052,"cache_read":0.04035},"canonical_model_id":"stepfun/step-3.7-flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.04613,"output":2.98895,"cache_read":0.10461},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.10461,"output":0.31384,"cache_read":0.02092},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"stepaudio-2.5-asr":{"id":"stepaudio-2.5-asr","type":"transcription","name":"StepAudio 2.5 ASR","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"step","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-24","last_updated":"2026-07-02","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}},"stepaudio-2.5-tts":{"id":"stepaudio-2.5-tts","type":"speech","name":"StepAudio 2.5 TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.10461,"output":0.31384,"cache_read":0.02092},"canonical_model_id":"stepfun/step-3.5-flash"},"step-tts-2":{"id":"step-tts-2","type":"speech","name":"Step TTS 2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-01","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}}}},"openreason":{"id":"openreason","env":["OPENREASON_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.openreason.app/v1","name":"OpenReason","doc":"https://openreason.app/docs","models":{"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.0022,"output":4.22},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-ai/deepseek-v4-flash-0731":{"id":"deepseek-ai/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.1371,"output":0.2743},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1055,"output":0.422},"canonical_model_id":"openai/gpt-oss-120b"}}},"umans-ai":{"id":"umans-ai","env":["UMANS_AI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.code.umans.ai/v1","name":"Umans AI","doc":"https://app.umans.ai/offers/code/docs/orgs","models":{"umans-glm-5.3":{"id":"umans-glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"umans-deepseek-v4.1-flash-1017":{"id":"umans-deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0.15,"output":0.6,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"umans-deepseek-v4.1-flash":{"id":"umans-deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0.15,"output":0.6,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"umans-coder":{"id":"umans-coder","name":"Umans Coder","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-glm-5.3-flash":{"id":"umans-glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-kimi-k3":{"id":"umans-kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"umans-flash":{"id":"umans-flash","name":"Umans Flash","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"umans-deepseek-v4-flash-0731":{"id":"umans-deepseek-v4-flash-0731","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"}}},"llmgateway":{"id":"llmgateway","env":["LLMGATEWAY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.llmgateway.io/v1","name":"DevPass (LLM Gateway)","doc":"https://llmgateway.io/docs","models":{"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"llama-4-scout-17b-instruct":{"id":"llama-4-scout-17b-instruct","name":"Llama 4 Scout 17B Instruct","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":2048},"cost":{"input":0.18,"output":0.59},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen3-vl-235b-a22b-instruct":{"id":"qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.737,"output":3.08,"cache_read":0.143},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.06,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"llama-3-70b-instruct":{"id":"llama-3-70b-instruct","name":"Llama 3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-04-18","last_updated":"2024-04-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8000},"cost":{"input":0.51,"output":0.74}},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.25,"cache_read":0.01},"canonical_model_id":"google/gemma-4-31b-it"},"mistral-small-2506":{"id":"mistral-small-2506","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/mistral-small-2506"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.032,"output":0.14,"cache_read":0.032},"canonical_model_id":"openai/gpt-oss-120b"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.135,"output":0.4},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.605,"output":2.123,"cache_read":0.154,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"minimax-m2.7-highspeed":{"id":"minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"sonar":{"id":"sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":130000,"output":4096},"cost":{"input":0.25,"output":2.5,"cache_write":0.25},"canonical_model_id":"perplexity/sonar"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"Ministral 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"mistral/ministral-14b-2512"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":512000},"cost":{"input":0.1,"output":0.5,"cache_read":0.005},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":1050000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"devstral-2512":{"id":"devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-2512"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.8-flash"},"qwen3-vl-30b-a3b-instruct":{"id":"qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen-plus":{"id":"qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.4,"output":1.2,"reasoning":4,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen-plus"},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65536},"cost":{"input":0.07,"output":0.27},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":384000},"cost":{"input":0.1,"output":0.3,"cache_read":0.0025},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"hy-mt2-plus":{"id":"hy-mt2-plus","name":"Hy-MT2 Plus","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"Hy","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.074,"output":0.295}},"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.45,"output":2,"cache_read":0.19,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"ling-3.0-flash-vl":{"id":"ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"qwen3-235b-a22b-fp8":{"id":"qwen3-235b-a22b-fp8","name":"Qwen3 235B A22B FP8","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":8192},"cost":{"input":0.2,"output":0.8}},"fugu-max":{"id":"fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":1050000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"qwen-max":{"id":"qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4},"canonical_model_id":"alibaba/qwen-max"},"grok-4-7":{"id":"grok-4-7","name":"Grok 4.7","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":384000},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.7-flash"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.04,"output":0.19,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-20b"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"seed-1-8-251228":{"id":"seed-1-8-251228","name":"Seed 1.8 (251228)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"ling-3.0-flash":{"id":"ling-3.0-flash","name":"InclusionAI Ling 3.0 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"minimax-text-01":{"id":"minimax-text-01","name":"MiniMax Text 01","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2,"output":1.1}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"grok-4-3":{"id":"grok-4-3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":65536},"cost":{"input":0.15,"output":1.35},"canonical_model_id":"alibaba/qwen3.6-27b"},"seed-1-6-250615":{"id":"seed-1-6-250615","name":"Seed 1.6 (250615)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-25","last_updated":"2025-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":1.6,"reasoning":4.8,"cache_read":0.04,"cache_write":0.25},"canonical_model_id":"alibaba/qwen3-vl-plus"},"gpt-5.2-pro":{"id":"gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"glm-4-32b-0414-128k":{"id":"glm-4-32b-0414-128k","name":"GLM-4 32B (0414-128k)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.1}},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.574,"output":3.011,"cache_read":0.225},"canonical_model_id":"moonshotai/kimi-k2.5"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"codestral-2508":{"id":"codestral-2508","name":"Codestral","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16384},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-2508"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"fugu-ultra-v2.0":{"id":"fugu-ultra-v2.0","name":"Fugu Ultra v2.0","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5}},"nemotron-3-ultra-550b":{"id":"nemotron-3-ultra-550b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":0.5,"output":2.2,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.108,"output":0.675,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3-coder-next"},"minimax-m2.1":{"id":"minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.27,"output":1.1},"canonical_model_id":"minimax/MiniMax-M2.1"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"grok-4-5":{"id":"grok-4-5","name":"Grok 4.5","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"seed-1-6-flash-250715":{"id":"seed-1-6-flash-250715","name":"Seed 1.6 Flash (250715)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.07,"output":0.3,"cache_read":0.015}},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gpt-live-transcribe":{"id":"gpt-live-transcribe","type":"transcription","name":"GPT-Live-Transcribe","description":"Low-latency streaming speech-to-text model with keyword and language hints","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-live-transcribe"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"grok-build-0-1":{"id":"grok-build-0-1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"glm-4.5-airx":{"id":"glm-4.5-airx","name":"GLM-4.5 AirX","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.1,"output":4.5,"cache_read":0.22}},"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8 27B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":32768},"cost":{"input":0.08,"output":0.35,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct (2507)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.09,"output":0.58},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.05,"cache_read":0.13},"canonical_model_id":"moonshotai/kimi-k2.6"},"ministral-8b-2512":{"id":"ministral-8b-2512","name":"Ministral 8B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.15,"output":0.15},"canonical_model_id":"mistral/ministral-8b-2512"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.08,"output":0.32,"cache_read":0.017,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"gemini-pro-latest":{"id":"gemini-pro-latest","name":"Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-27","last_updated":"2026-02-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2}},"grok-4":{"id":"grok-4","name":"Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":3,"output":15,"cache_read":0.75}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.12,"output":0.25,"cache_read":0.0025},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"qwen3.5-122b-a10b":{"id":"qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"qwen3-14b":{"id":"qwen3-14b","name":"Qwen3 14B","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.076,"output":0.228}},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"glm-5.2-fast":{"id":"glm-5.2-fast","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.156,"output":6.776,"cache_read":0.429},"canonical_model_id":"zhipuai/glm-5.2"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"gpt-4o-transcribe":{"id":"gpt-4o-transcribe","name":"GPT-4o Transcribe","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":2.5,"output":10}},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"grok-4-20-beta-0309-reasoning":{"id":"grok-4-20-beta-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"muse-spark-1.2-contributor":{"id":"muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.05,"cache_write":0.3125},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.845,"output":3.38,"cache_read":0.6,"cache_write":3.75},"canonical_model_id":"alibaba/qwen3-max"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"cost":{"input":0.573,"output":2.58,"cache_read":0.144,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"kimi-k3-fast":{"id":"kimi-k3-fast","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1040384,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"glm-4.7-flashx":{"id":"glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.45,"output":1.8,"cache_read":0.135},"canonical_model_id":"zhipuai/glm-5.3"},"llama-3.1-70b-instruct":{"id":"llama-3.1-70b-instruct","name":"Llama 3.1 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"status":"beta","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"granite-4.2-8b":{"id":"granite-4.2-8b","name":"Granite 4.2 8B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"llama-4-maverick-17b-instruct":{"id":"llama-4-maverick-17b-instruct","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":2048},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.36,"output":0.87,"reasoning":8.4},"canonical_model_id":"alibaba/qwen3-32b"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":3.75,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"minimax-m2.1-lightning":{"id":"minimax-m2.1-lightning","name":"MiniMax M2.1 Lightning","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.12,"output":0.48}},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"qwen-flash":{"id":"qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.05,"output":0.4,"cache_read":0.01,"cache_write":0.0625},"canonical_model_id":"alibaba/qwen-flash"},"gpt-transcribe":{"id":"gpt-transcribe","type":"transcription","name":"GPT-Transcribe","description":"High-accuracy speech-to-text model for audio files and committed Realtime turns","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-transcribe"},"minimax-m2.5-highspeed":{"id":"minimax-m2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"qwen3-235b-a22b-thinking-2507":{"id":"qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22B Thinking (2507)","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":8192},"cost":{"input":0.3,"output":3}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"mistral-large-latest":{"id":"mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"cost":{"input":4,"output":12},"canonical_model_id":"mistral/mistral-large-latest"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":228700,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"glm-4.5-x":{"id":"glm-4.5-x","name":"GLM-4.5 X","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"beta","cost":{"input":2.2,"output":8.9,"cache_read":0.45}},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"seed-2-0-mini-260428":{"id":"seed-2-0-mini-260428","name":"Seed 2.0 Mini (260428)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.4,"cache_read":0.02}},"seed-2-0-code-preview-260328":{"id":"seed-2-0-code-preview-260328","name":"Seed 2.0 Code Preview (260328)","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"glm-5.3-fast":{"id":"glm-5.3-fast","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.56},"canonical_model_id":"zhipuai/glm-5.3"},"llama-3.2-3b-instruct":{"id":"llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-09-18","last_updated":"2024-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.03,"output":0.05}},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"fugu-ultra":{"id":"fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"seed-2-0-lite-260428":{"id":"seed-2-0-lite-260428","name":"Seed 2.0 Lite (260428)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-2603"},"seed-1-6-250915":{"id":"seed-1-6-250915","name":"Seed 1.6 (250915)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"glm-4.6v-flashx":{"id":"glm-4.6v-flashx","name":"GLM-4.6V FlashX","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.04,"output":0.4,"cache_read":0.004}},"glm-5-turbo":{"id":"glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"auto":{"id":"auto","name":"Auto Route","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0,"output":0}},"gpt-4o-mini-transcribe":{"id":"gpt-4o-mini-transcribe","name":"GPT-4o Mini Transcribe","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":1.25,"output":5}},"kimi-k3.1-pro-highspeed":{"id":"kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4}},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.54,"output":4.631,"cache_read":0.198,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":384000},"cost":{"input":0.076,"output":0.153,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.088,"output":0.25,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"grok-4-20-reasoning":{"id":"grok-4-20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"seed-2-0-pro-260328":{"id":"seed-2-0-pro-260328","name":"Seed 2.0 Pro (260328)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"grok-4-20-non-reasoning":{"id":"grok-4-20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.095,"output":0.285},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"kimi-k2":{"id":"kimi-k2","name":"Kimi K2","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.57,"output":2.3,"cache_read":0.5}},"qwen3-vl-235b-a22b-thinking":{"id":"qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"ernie-4.5-vl-424b-a47b":{"id":"ernie-4.5-vl-424b-a47b","name":"ERNIE 4.5 VL 424B A47B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":123000,"output":123000},"cost":{"input":0.42,"output":1.25}},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen-plus-latest":{"id":"qwen-plus-latest","name":"Qwen Plus Latest","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-01-25","last_updated":"2025-01-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":8192},"cost":{"input":0.4,"output":1.2,"cache_read":0.08,"cache_write":0.5}},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.1-flash-lite"},"grok-4-20-beta-0309-non-reasoning":{"id":"grok-4-20-beta-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"custom":{"id":"custom","name":"Custom Model","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0,"output":0}},"mistral-medium-3-5":{"id":"mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"qwen35-397b-a17b":{"id":"qwen35-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen3-coder-flash":{"id":"qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":997952,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":1000000},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.0375},"canonical_model_id":"alibaba/qwen3.7-flash"},"ministral-3b-2512":{"id":"ministral-3b-2512","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"mistral/ministral-3b-2512"},"minimax-m2":{"id":"minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.2,"output":1,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2"},"llama-3.2-11b-instruct":{"id":"llama-3.2-11b-instruct","name":"Llama 3.2 11B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.07,"output":0.33}},"reka-edge-2603":{"id":"reka-edge-2603","name":"Reka Edge 2603","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-09","last_updated":"2026-10-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":16384},"cost":{"input":0.1,"output":0.1}}}},"bothub":{"id":"bothub","env":["BOTHUB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://openai.bothub.ru/v1","name":"Bothub","doc":"https://bothub.ru/models","models":{"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"meta/muse-spark-1.3"},"nemotron-3-ultra-550b-a55b:free":{"id":"nemotron-3-ultra-550b-a55b:free","name":"Nemotron 3 Ultra (free)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gemma-4-31b-it:free":{"id":"gemma-4-31b-it:free","name":"Gemma 4 31B IT (free)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.7671,"output":5.3011},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.589,"output":1.7671},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4243,"output":4.4765},"canonical_model_id":"zhipuai/glm-5.3"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2677,"output":1.6064},"canonical_model_id":"openai/gpt-5.6-luna"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1566,"output":0.522},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"huggingface":{"id":"huggingface","env":["HF_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://router.huggingface.co/v1","name":"Hugging Face","doc":"https://huggingface.co/docs/inference-providers","models":{"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"MiniMaxAI/MiniMax-M1-80k":{"id":"MiniMaxAI/MiniMax-M1-80k","name":"MiniMax M1 80K","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":80000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"minimax/minimax-m1-80k"},"MiniMaxAI/MiniMax-M2.1":{"id":"MiniMaxAI/MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-10","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"MiniMaxAI/MiniMax-M2":{"id":"MiniMaxAI/MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":4.05},"canonical_model_id":"thinkingmachines/inkling"},"tencent/Hy4-preview":{"id":"tencent/Hy4-preview","name":"Hy4 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501},"canonical_model_id":"tencent/hy4-preview"},"tencent/Hy3":{"id":"tencent/Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.14,"output":0.58},"canonical_model_id":"tencent/hy3"},"meta-llama/Llama-Guard-4-12B":{"id":"meta-llama/Llama-Guard-4-12B","name":"Llama Guard 4 12B","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/Llama-3.1-8B-Instruct":{"id":"meta-llama/Llama-3.1-8B-Instruct","name":"Llama-3.1-8B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.06,"output":0.06},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2-Instruct":{"id":"moonshotai/Kimi-K2-Instruct","name":"Kimi-K2-Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1,"output":3},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/Kimi-K2-Instruct-0905":{"id":"moonshotai/Kimi-K2-Instruct-0905","name":"Kimi-K2-Instruct-0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":1,"output":3},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2-Thinking":{"id":"moonshotai/Kimi-K2-Thinking","name":"Kimi-K2-Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen3 8B","description":"Dense 8B Qwen3 model with switchable thinking for chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":32768},"cost":{"input":0.07,"output":0.18},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-Embedding-4B":{"id":"Qwen/Qwen3-Embedding-4B","type":"embedding","name":"Qwen 3 Embedding 4B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-12","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://router.huggingface.co/deepinfra/v1/openai"},"cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-4b"},"Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen3-Coder-480B-A35B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":2,"output":2},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3-VL-235B-A22B-Instruct":{"id":"Qwen/Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.855,"output":2.565},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-30B-A3B":{"id":"Qwen/Qwen3-30B-A3B","name":"Qwen3 30B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3-235B-A22B-Thinking-2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.3,"output":3},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.29,"output":0.59},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.47,"output":3.19},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":3},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":6.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next-80B-A3B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.25,"output":1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"Qwen/Qwen2.5-Coder-32B-Instruct":{"id":"Qwen/Qwen2.5-Coder-32B-Instruct","name":"Qwen2.5-Coder-32B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.06,"output":0.2},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"Qwen/Qwen3-Coder-Next":{"id":"Qwen/Qwen3-Coder-Next","name":"Qwen3-Coder-Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"Qwen/Qwen3-VL-235B-A22B-Thinking":{"id":"Qwen/Qwen3-VL-235B-A22B-Thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.07,"output":0.26},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.15,"output":0.95},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5-397B-A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3-235B-A22B":{"id":"Qwen/Qwen3-235B-A22B","name":"Qwen3 235B-A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"Qwen/Qwen3-Embedding-8B":{"id":"Qwen/Qwen3-Embedding-8B","type":"embedding","name":"Qwen 3 Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-12","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://router.huggingface.co/deepinfra/v1/openai"},"cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.27,"output":1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.27,"output":1.12},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"id":"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":32768},"cost":{"input":0.7,"output":2.5},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":8192},"cost":{"input":0.4,"output":1.3},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.28,"output":0.4},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek-R1-0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":3,"output":5},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.25,"output":0.69},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-oss-20b"},"stepfun-ai/Step-3.7-Flash":{"id":"stepfun-ai/Step-3.7-Flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun-ai/Step-3.5-Flash":{"id":"stepfun-ai/Step-3.5-Flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"XiaomiMiMo/MiMo-V2.5-Pro":{"id":"XiaomiMiMo/MiMo-V2.5-Pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":3},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"XiaomiMiMo/MiMo-V2.5":{"id":"XiaomiMiMo/MiMo-V2.5","name":"MiMo-V2.5","description":"MiMo model for long-context reasoning, perception, and agentic tasks","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.4,"output":2},"canonical_model_id":"xiaomi/mimo-v2.5"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai-org/AutoGLM-Phone-9B-Multilingual":{"id":"zai-org/AutoGLM-Phone-9B-Multilingual","name":"AutoGLM Phone 9B Multilingual","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-12-10","last_updated":"2025-12-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.035,"output":0.138},"canonical_model_id":"zhipuai/autoglm-phone-9b-multilingual"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-4.5V":{"id":"zai-org/GLM-4.5V","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8},"canonical_model_id":"zhipuai/glm-4.5v"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-4.7-Flash":{"id":"zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/GLM-4.6V-Flash":{"id":"zai-org/GLM-4.6V-Flash","name":"GLM-4.6V-Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":1.4,"output":4.4},"canonical_model_id":"zhipuai/glm-5.2"}}},"ainetcafe":{"id":"ainetcafe","env":["AINETCAFE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://microquickjs.com/v1","name":"ainetcafe","doc":"https://ainetcafe.com/k3/guides/","models":{"Kimi-K3":{"id":"Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":2.1,"output":10.5,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"}}},"ovhcloud":{"id":"ovhcloud","env":["OVHCLOUD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://oai.endpoints.kepler.ai.cloud.ovh.net/v1","name":"OVHcloud AI Endpoints","doc":"https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog//","models":{"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5-9B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.12,"output":0.18},"canonical_model_id":"alibaba/qwen3.5-9b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"gpt-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.09,"output":0.47},"canonical_model_id":"openai/gpt-oss-120b"},"qwen2.5-vl-72b-instruct":{"id":"qwen2.5-vl-72b-instruct","name":"Qwen2.5-VL-72B-Instruct","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":1.01,"output":1.01},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"gpt-oss-20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6-27B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.47,"output":3.19},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen3guard-gen-8b":{"id":"qwen3guard-gen-8b","name":"Qwen3Guard-Gen-8B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-22","last_updated":"2026-01-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0,"output":0}},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5-397B-A17B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.71,"output":4.25},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"meta-llama-3_3-70b-instruct":{"id":"meta-llama-3_3-70b-instruct","name":"Meta-Llama-3_3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.74,"output":0.74},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"mistral-nemo-instruct-2407":{"id":"mistral-nemo-instruct-2407","name":"Mistral-Nemo-Instruct-2407","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.14,"output":0.14},"canonical_model_id":"mistral/mistral-nemo"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8-27B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.47,"output":3.19},"canonical_model_id":"alibaba/qwen3.8-27b"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"Mistral-Small-3.2-24B-Instruct-2506","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.31},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"qwen3guard-gen-0.6b":{"id":"qwen3guard-gen-0.6b","name":"Qwen3Guard-Gen-0.6B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-22","last_updated":"2026-01-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0,"output":0}}}},"xiaomi-token-plan-ams":{"id":"xiaomi-token-plan-ams","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan-ams.xiaomimimo.com/v1","name":"Xiaomi Token Plan (Europe)","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.5-tts":{"id":"mimo-v2.5-tts","type":"speech","name":"MiMo-V2.5-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5-tts-voiceclone":{"id":"mimo-v2.5-tts-voiceclone","type":"speech","name":"MiMo-V2.5-TTS-VoiceClone","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voiceclone"},"mimo-v2.5-tts-voicedesign":{"id":"mimo-v2.5-tts-voicedesign","type":"speech","name":"MiMo-V2.5-TTS-VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voicedesign"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"}}},"minimax-cn":{"id":"minimax-cn","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.cn/anthropic/v1","name":"MiniMax (minimax.cn)","doc":"https://platform.minimaxi.com/docs/guides/quickstart","models":{"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6275,"output":2.5098,"cache_read":0.0627,"cache_write":0.3922}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.1494,"output":0.5976,"cache_read":0.0149},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0314,"cache_write":0.3922}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0314,"cache_write":0.3922}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0.3137,"output":1.2549,"cache_read":0.0627,"tiers":[{"input":0.6275,"output":2.5098,"cache_read":0.1255,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6275,"output":2.5098,"cache_read":0.1255}}},"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6275,"output":2.5098,"cache_read":0.0314,"cache_write":0.3922}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0627,"cache_write":0.3922}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3137,"output":1.2549,"cache_read":0.0314,"cache_write":0.3922}}}},"xpersona":{"id":"xpersona","env":["XPERSONA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://www.xpersona.co/v1","name":"Xpersona","doc":"https://www.xpersona.co/docs","models":{"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.55,"output":12.2,"reasoning":12.2,"cache_read":0.155},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.75,"output":6,"reasoning":6,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":1.5,"output":12,"reasoning":12,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.9,"output":5.55,"reasoning":5.55,"cache_read":0.09},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":3.7,"reasoning":3.7,"cache_read":0.06},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":18.5,"reasoning":18.5,"cache_read":0.3},"canonical_model_id":"anthropic/claude-fable-5"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":0.375,"output":4,"reasoning":4,"cache_read":0.0375},"canonical_model_id":"openai/gpt-5.4-mini"},"xpersona-frieren-coder":{"id":"xpersona-frieren-coder","name":"Xpersona Frieren 1","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-30","release_date":"2026-05-01","last_updated":"2026-05-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":384000},"cost":{"input":1.5,"output":6,"reasoning":6,"cache_read":0.15}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":372000,"output":128000},"cost":{"input":1.5,"output":2,"reasoning":2,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.6-terra"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.5,"output":9.25,"reasoning":9.25,"cache_read":0.15},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.6":{"id":"gpt-5.6","name":"GPT-5.6","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":372000,"output":128000},"cost":{"input":1.5,"output":12,"reasoning":12,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.6-sol"},"xpersona-gpt-5.5":{"id":"xpersona-gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-30","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":3,"output":18,"reasoning":18,"cache_read":0.3},"canonical_model_id":"openai/gpt-5.5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":372000,"output":128000},"cost":{"input":1.5,"output":12,"reasoning":12,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.6-sol"}}},"clarifai":{"id":"clarifai","env":["CLARIFAI_PAT"],"npm":"@ai-sdk/openai-compatible","api":"https://api.clarifai.com/v2/ext/openai/v1","name":"Clarifai","doc":"https://docs.clarifai.com/compute/inference/","models":{"mistralai/completion/models/Ministral-3-14B-Reasoning-2512":{"id":"mistralai/completion/models/Ministral-3-14B-Reasoning-2512","name":"Ministral 3 14B Reasoning 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-01","last_updated":"2025-12-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2.5,"output":1.7}},"mistralai/completion/models/Ministral-3-3B-Reasoning-2512":{"id":"mistralai/completion/models/Ministral-3-3B-Reasoning-2512","name":"Ministral 3 3B Reasoning 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12","last_updated":"2026-02-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.039,"output":0.54825}},"clarifai/main/models/mm-poly-8b":{"id":"clarifai/main/models/mm-poly-8b","name":"MM Poly 8B","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"mm-poly","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-06","last_updated":"2026-02-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"cost":{"input":0.658,"output":1.11}},"moonshotai/chat-completion/models/Kimi-K2_6":{"id":"moonshotai/chat-completion/models/Kimi-K2_6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"arcee_ai/AFM/models/trinity-mini":{"id":"arcee_ai/AFM/models/trinity-mini","name":"Trinity Mini","description":"Reasoning-tuned 26B MoE model with 3B active parameters for agents, tools, and multi-step workloads","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-12-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.045,"output":0.15},"canonical_model_id":"arcee-ai/trinity-mini"},"deepseek-ai/deepseek-ocr/models/DeepSeek-OCR":{"id":"deepseek-ai/deepseek-ocr/models/DeepSeek-OCR","name":"DeepSeek OCR","description":"OCR model for extracting structured text from documents and screenshots","family":"deepseek","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2026-02-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.2,"output":0.7}},"openai/chat-completion/models/gpt-oss-20b":{"id":"openai/chat-completion/models/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.045,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"openai/chat-completion/models/gpt-oss-120b-high-throughput":{"id":"openai/chat-completion/models/gpt-oss-120b-high-throughput","name":"GPT OSS 120B High Throughput","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.09,"output":0.36},"canonical_model_id":"openai/gpt-oss-120b"},"qwen/qwenLM/models/Qwen3-30B-A3B-Instruct-2507":{"id":"qwen/qwenLM/models/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwenLM/models/Qwen3-30B-A3B-Thinking-2507":{"id":"qwen/qwenLM/models/Qwen3-30B-A3B-Thinking-2507","name":"Qwen3 30B A3B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-31","last_updated":"2026-02-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.36,"output":1.3}},"qwen/qwenCoder/models/Qwen3-Coder-30B-A3B-Instruct":{"id":"qwen/qwenCoder/models/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.11458,"output":0.74812},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"minimaxai/chat-completion/models/MiniMax-M2_5-high-throughput":{"id":"minimaxai/chat-completion/models/MiniMax-M2_5-high-throughput","name":"MiniMax-M2.5 High Throughput","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"tokenrouter":{"id":"tokenrouter","env":["TOKENROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tokenrouter.com/v1","name":"TokenRouter","doc":"https://www.tokenrouter.com/docs/tokenrouter-feature-guide/","models":{"z-ai/glm-5.3-free":{"id":"z-ai/glm-5.3-free","name":"GLM-5.3 (free)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"}}},"thinkingmachines":{"id":"thinkingmachines","env":["TINKER_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://tinker.thinkingmachines.dev/services/tinker-prod/anthropic/api/v1","name":"Thinking Machines","doc":"https://tinker-docs.thinkingmachines.ai/tinker/compatible-apis/anthropic/","models":{"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.184,"output":0.495,"cache_read":0.0368},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16:peft:262144","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.184,"output":0.495,"cache_read":0.0368},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16:peft:262144","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.195,"output":0.495,"cache_read":0.039},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16:peft:262144","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.8,"output":6.225,"cache_read":0.36},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.8,"output":6.225,"cache_read":0.36},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16:peft:262144":{"id":"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16:peft:262144","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57,"output":1.44,"cache_read":0.114},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.58,"output":1.44,"cache_read":0.116},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/Inkling:peft:262144":{"id":"thinkingmachines/Inkling:peft:262144","name":"Inkling (256K)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling:peft:262144:sampling-nvfp4":{"id":"thinkingmachines/Inkling:peft:262144:sampling-nvfp4","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling-Small:peft:262144":{"id":"thinkingmachines/Inkling-Small:peft:262144","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.58,"output":1.44,"cache_read":0.116},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/Inkling-Small:peft:262144:sampling-nvfp4":{"id":"thinkingmachines/Inkling-Small:peft:262144:sampling-nvfp4","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"thinkingmachines/inkling-small"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":2.205,"output":5.49,"cache_read":0.441},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.6:peft:131072":{"id":"moonshotai/Kimi-K2.6:peft:131072","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":2.205,"output":5.49,"cache_read":0.441},"canonical_model_id":"moonshotai/kimi-k2.6"},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen3 8B","description":"Dense 8B Qwen3 model with switchable thinking for chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.195,"output":0.6,"cache_read":0.039},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3.6-35B-A3B:peft:262144":{"id":"Qwen/Qwen3.6-35B-A3B:peft:262144","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":1.335,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.5-397B-A17B:peft:262144":{"id":"Qwen/Qwen3.5-397B-A17B:peft:262144","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.65,"output":7.5,"cache_read":0.33},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.86,"output":5.595,"cache_read":0.372},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3.5-4B":{"id":"Qwen/Qwen3.5-4B","name":"Qwen3.5 4B","description":"Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.33,"output":1.005,"cache_read":0.066},"canonical_model_id":"alibaba/qwen3.5-4b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.6975,"output":5.595,"cache_read":0.1395},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.5-9B-Base":{"id":"Qwen/Qwen3.5-9B-Base","name":"Qwen3.5 9B Base","description":"Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.3,"output":0.807,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3.5-9b-base"},"Qwen/Qwen3.5-4B:peft:262144":{"id":"Qwen/Qwen3.5-4B:peft:262144","name":"Qwen3.5 4B","description":"Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.33,"output":1.005,"cache_read":0.066},"canonical_model_id":"alibaba/qwen3.5-4b"},"Qwen/Qwen3.8-27B:peft:262144":{"id":"Qwen/Qwen3.8-27B:peft:262144","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6975,"output":5.595,"cache_read":0.1395},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.45,"output":1.335,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3.5-35B-A3B-Base":{"id":"Qwen/Qwen3.5-35B-A3B-Base","name":"Qwen3.5 35B-A3B Base","description":"Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.54,"output":1.335,"cache_read":0.108},"canonical_model_id":"alibaba/qwen3.5-35b-a3b-base"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.3,"output":0.807,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.65,"output":7.5,"cache_read":0.33},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":1.695,"output":4.215,"cache_read":0.339},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V4.1-Flash-1017:peft:262144":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017:peft:262144","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.6,"cache_read":0.12},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-V4.1-Flash:peft:262144":{"id":"deepseek-ai/DeepSeek-V4.1-Flash:peft:262144","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.6,"cache_read":0.12},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.33,"output":0.84,"cache_read":0.066},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.18,"output":0.45,"cache_read":0.036},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-120b:peft:131072":{"id":"openai/gpt-oss-120b:peft:131072","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.33,"output":0.84,"cache_read":0.066},"canonical_model_id":"openai/gpt-oss-120b"},"zai-org/GLM-5.3:peft:262144":{"id":"zai-org/GLM-5.3:peft:262144","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":3.78,"output":12.15,"cache_read":0.756},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash:peft:262144":{"id":"zai-org/GLM-5.3-Flash:peft:262144","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.84,"output":2.11,"cache_read":0.168},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"alibaba-coding-plan-cn":{"id":"alibaba-coding-plan-cn","env":["ALIBABA_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://coding.dashscope.aliyuncs.com/v1","name":"Alibaba Coding Plan (China)","doc":"https://help.aliyun.com/zh/model-studio/coding-plan","models":{"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-11","last_updated":"2026-02-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-max-2026-01-23":{"id":"qwen3-max-2026-01-23","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":24576},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}}}},"ai21":{"id":"ai21","env":["AI21_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ai21.com/studio/v1","name":"AI21 Labs","doc":"https://docs.ai21.com/docs/jamba-foundation-models","models":{"jamba-mini":{"id":"jamba-mini","name":"Jamba Mini","description":"AI21's efficient, lightweight hybrid SSM-Transformer model for a wide range of tasks","family":"jamba","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-22","release_date":"2026-01-08","last_updated":"2026-01-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"ai21/jamba-mini"},"jamba-large":{"id":"jamba-large","name":"Jamba Large","description":"AI21's hybrid SSM-Transformer long-context model for enterprise agents and grounded generation","family":"jamba","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-22","release_date":"2025-07-03","last_updated":"2025-07-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"ai21/jamba-large"}}},"hetzner":{"id":"hetzner","env":["HETZNER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.hetzner.com/api/v1","name":"Hetzner","doc":"https://experiments.hetzner.com/docs/inference","models":{"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8-27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"beta","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.6-35B-A3B-FP8":{"id":"Qwen/Qwen3.6-35B-A3B-FP8","name":"Qwen3.6 35B A3B FP8","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"beta","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"atomic-chat":{"id":"atomic-chat","env":["ATOMIC_CHAT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://127.0.0.1:1337/v1","name":"Atomic Chat","doc":"https://atomic.chat","models":{"gemma-4-E4B-it-MLX-4bit":{"id":"gemma-4-E4B-it-MLX-4bit","name":"Gemma 4 E4B Instruct (MLX 4-bit)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-E4B-it"},"gemma-4-E4B-it-IQ4_XS":{"id":"gemma-4-E4B-it-IQ4_XS","name":"Gemma 4 E4B Instruct (IQ4_XS)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-E4B-it"},"Meta-Llama-3_1-8B-Instruct-GGUF":{"id":"Meta-Llama-3_1-8B-Instruct-GGUF","name":"Meta Llama 3.1 8B Instruct (GGUF)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"Qwen3_5-9B-Q4_K_M":{"id":"Qwen3_5-9B-Q4_K_M","name":"Qwen 3.5 9B (Q4_K_M)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen3_5-9B-MLX-4bit":{"id":"Qwen3_5-9B-MLX-4bit","name":"Qwen 3.5 9B (MLX 4-bit)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-9b"}}},"stepfun-ai-step-plan":{"id":"stepfun-ai-step-plan","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.ai/step_plan/v1","name":"StepFun Step Plan (Global)","doc":"https://platform.stepfun.ai/docs/en/step-plan/integrations/reasoning-api","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.7-flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash"}}},"crusoe":{"id":"crusoe","env":["CRUSOE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inference.crusoecloud.com/v1","name":"Crusoe","doc":"https://docs.crusoecloud.com/managed-inference/overview","models":{"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4,"cache_read":0.14},"canonical_model_id":"google/gemma-4-31b-it"},"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B":{"id":"nvidia/NVIDIA-Nemotron-3-Super-120B-A12B","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.4,"cache_read":0.15},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B":{"id":"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.03},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/Nemotron-3-Nano-Omni-Reasoning-30B-A3B":{"id":"nvidia/Nemotron-3-Nano-Omni-Reasoning-30B-A3B","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.3,"output":1.83,"cache_read":0.3,"input_audio":0.5},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.25,"output":0.75,"cache_read":0.13},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.7,"output":3.5,"cache_read":0.35},"canonical_model_id":"moonshotai/kimi-k2.6"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.22,"output":0.8,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"status":"deprecated","cost":{"input":0.5,"output":1.5,"cache_read":0.25},"canonical_model_id":"deepseek/deepseek-v3-0324"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.05},"canonical_model_id":"openai/gpt-oss-120b"},"zai/GLM-5.1":{"id":"zai/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4.4,"cache_read":0.25},"canonical_model_id":"zhipuai/glm-5.1"},"zai/GLM-5.2":{"id":"zai/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"xai":{"id":"xai","env":["XAI_API_KEY"],"npm":"@ai-sdk/xai","name":"xAI","doc":"https://docs.x.ai/docs/models","models":{"grok-4.20-0309-reasoning":{"id":"grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"grok-imagine-image":{"id":"grok-imagine-image","type":"image","name":"Grok Imagine Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"canonical_model_id":"xai/grok-imagine-image"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"grok-imagine-video":{"id":"grok-imagine-video","type":"video","name":"Grok Imagine Video","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0}},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-imagine-image-quality":{"id":"grok-imagine-image-quality","type":"image","name":"Grok Imagine Image Quality","description":"Higher-fidelity Grok Imagine image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-03","last_updated":"2026-04-03","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"status":"deprecated","canonical_model_id":"xai/grok-imagine-image-quality"},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}}},"grok-4.20-0309-non-reasoning":{"id":"grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-imagine-video-1.5-lite":{"id":"grok-imagine-video-1.5-lite","type":"video","name":"Grok Imagine Video 1.5 Lite","description":"Cost-efficient video model for text-to-video and image-to-video generation","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"canonical_model_id":"xai/grok-imagine-video-1.5-lite"},"grok-4.20-multi-agent-0309":{"id":"grok-4.20-multi-agent-0309","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"grok-imagine-video-1.5":{"id":"grok-imagine-video-1.5","type":"video","name":"Grok Imagine Video 1.5","description":"Video model for text-to-video, image-to-video, and reference-to-video generation with optional voice references","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-30","last_updated":"2026-05-30","modalities":{"input":["text","image","audio"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"canonical_model_id":"xai/grok-imagine-video-1.5"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"}}},"amd":{"id":"amd","env":["AMD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://developer.amd.com.cn/radeon/api/v1","name":"AMD","doc":"https://developer.amd.com.cn/radeon/tokenfactory","models":{"Gemma-4-31B-IT":{"id":"Gemma-4-31B-IT","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"google/gemma-4-31b-it"},"Qwen3.8-Flash-Next":{"id":"Qwen3.8-Flash-Next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"MiniCPM5-2B":{"id":"MiniCPM5-2B","name":"MiniCPM5-2B","description":"Dense 2B-class open-source model for on-device and resource-constrained use, with native long-context support, tool calling, and agentic tasks","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-09-06","last_updated":"2026-09-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"openbmb/minicpm5-2b"},"MiniCPM-V-4.7-1B":{"id":"MiniCPM-V-4.7-1B","name":"MiniCPM-V 4.7 1B","description":"Compact MiniCPM-V model for single-image, multi-image and video understanding","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"openbmb/minicpm-v-4.7-1b"},"MiniCPM-V-4.7-35B-A3B":{"id":"MiniCPM-V-4.7-35B-A3B","name":"MiniCPM-V 4.7 35B-A3B","description":"First MoE MiniCPM-V model for single-image, multi-image and video understanding","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"openbmb/minicpm-v-4.7-35b-a3b"},"DeepSeek-V4.1-Flash":{"id":"DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"DeepSeek-V4-Flash-Vision-Exp":{"id":"DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-27b"},"MiMo-V2.6-Flash":{"id":"MiMo-V2.6-Flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"GLM-5.3-Flash":{"id":"GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"dinference":{"id":"dinference","env":["DINFERENCE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.dinference.com/v1","name":"DInference","doc":"https://dinference.com","models":{"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.0675,"output":0.27},"canonical_model_id":"openai/gpt-oss-120b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":3.89},"canonical_model_id":"zhipuai/glm-5.2"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.45,"output":1.65},"canonical_model_id":"zhipuai/glm-4.7"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":3.89},"canonical_model_id":"zhipuai/glm-5.1"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.75,"output":2.4},"canonical_model_id":"zhipuai/glm-5"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":0.22,"output":0.88},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"inference":{"id":"inference","env":["INFERENCE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.net/v1","name":"Inference","doc":"https://inference.net/models","models":{"google/gemma-3":{"id":"google/gemma-3","name":"Google Gemma 3","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":125000,"output":4096},"cost":{"input":0.15,"output":0.3}},"meta/llama-3.1-8b-instruct":{"id":"meta/llama-3.1-8b-instruct","name":"Llama 3.1 8B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.025,"output":0.025},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta/llama-3.2-1b-instruct":{"id":"meta/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"meta/llama-3.2-3b-instruct":{"id":"meta/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"cost":{"input":0.02,"output":0.02}},"meta/llama-3.2-11b-vision-instruct":{"id":"meta/llama-3.2-11b-vision-instruct","name":"Llama 3.2 11B Vision Instruct","description":"Open Llama multimodal model for image understanding and text reasoning","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.055,"output":0.055},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"osmosis/osmosis-structure-0.6b":{"id":"osmosis/osmosis-structure-0.6b","name":"Osmosis Structure 0.6B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"osmosis","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4000,"output":2048},"status":"deprecated","cost":{"input":0.1,"output":0.5}},"mistral/mistral-nemo-12b-instruct":{"id":"mistral/mistral-nemo-12b-instruct","name":"Mistral Nemo 12B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.038,"output":0.1},"canonical_model_id":"mistral/mistral-nemo"},"qwen/qwen-2.5-7b-vision-instruct":{"id":"qwen/qwen-2.5-7b-vision-instruct","name":"Qwen 2.5 7B Vision Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":125000,"output":4096},"status":"deprecated","cost":{"input":0.2,"output":0.2}},"qwen/qwen3-embedding-4b":{"id":"qwen/qwen3-embedding-4b","type":"embedding","name":"Qwen 3 Embedding 4B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-12","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":2048},"status":"deprecated","cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-4b"}}},"prime":{"id":"prime","env":["PRIME_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.pinference.ai/api/v1","name":"Prime Inference","doc":"https://docs.primeintellect.ai/inference/overview","models":{"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2.2,"output":6.6},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.165,"output":0.66},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral Nemo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.165,"output":0.17},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3 2512","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.54,"output":4.84},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.98,"output":3.08,"cache_read":0.182,"cache_write":0.98},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0.15},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.119,"output":0.3},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["audio","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["audio","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.35,"output":6.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-3.6-flash"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1.04,"output":2.253},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro 0423","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.91,"output":3.83},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash 0423","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.44,"output":1.32},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":3.45,"output":17.25},"canonical_model_id":"moonshotai/kimi-k3"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25},"canonical_model_id":"meta/muse-spark-open"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna Pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":1,"output":6},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.25,"output":2},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.05,"output":0.4},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.75,"output":4.5},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra Pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.2,"output":1.25},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.2,"output":1.2},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o-mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.13,"output":0.52},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":235929},"cost":{"input":0.99,"output":3.2},"canonical_model_id":"alibaba/qwen3.8-27b"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":1.6,"output":4.8},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6},"canonical_model_id":"xai/grok-4.6"}}},"orcarouter":{"id":"orcarouter","env":["ORCAROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.orcarouter.ai/v1","name":"OrcaRouter","doc":"https://docs.orcarouter.ai","models":{"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.26,"output":3.96,"cache_read":0.234},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5.3-flash-free":{"id":"z-ai/glm-5.3-flash-free","name":"GLM-5.3-Flash (free)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.13,"output":0.38,"cache_read":0.02},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-flash-lite-latest"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.08333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":18,"cache_read":0.4},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-robotics-er-1.6-preview":{"id":"google/gemini-robotics-er-1.6-preview","name":"Gemini Robotics-ER 1.6 Preview","description":"Vision-language model for embodied reasoning: spatial understanding, task planning, and physical-world agentic robotics","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-14","last_updated":"2026-04-14","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":1,"output":5},"canonical_model_id":"google/gemini-robotics-er-1.6-preview"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"google/gemini-flash-latest"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.33,"cache_read":0.0075},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"orcarouter/fusion-flash":{"id":"orcarouter/fusion-flash","name":"OrcaRouter Fusion Flash","description":"Budget Fusion panel that runs Gemini 3.5 Flash, MiniMax M2.7 and GLM 5.1 in parallel on hard requests, then has a Claude Opus 4.8 judge return the strongest single answer verbatim. Cost-sensitive fan-out over a 200K window.","family":"model-router","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000}},"orcarouter/free":{"id":"orcarouter/free","name":"OrcaRouter Free","description":"Built-in router over the free tier that scores each request's difficulty and sends light work to the smaller free model and harder work to the stronger one. Priced at zero and never falls back to a paid model.","family":"auto","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":0,"output":0}},"orcarouter/fusion":{"id":"orcarouter/fusion","name":"OrcaRouter Fusion","description":"Curated fan-out router that runs Claude Opus 4.8, GPT-5.5 and Gemini 3.1 Pro in parallel on hard requests, then has a Claude Opus 4.8 judge return the strongest single answer verbatim. Easy requests fall through to a cheaper default and bill as one call.","family":"model-router","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"orcarouter/fusion-mini":{"id":"orcarouter/fusion-mini","name":"OrcaRouter Fusion Mini","description":"Leaner two-model Fusion panel that runs Claude Opus 4.8 and GPT-5.5 in parallel on hard requests, then has a Claude Opus 4.8 judge return the strongest single answer verbatim. Easy requests fall through to a cheaper default and bill as one call.","family":"model-router","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"orcarouter/auto":{"id":"orcarouter/auto","name":"OrcaRouter Auto","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-01-01","last_updated":"2026-05-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0,"output":0}},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.18,"output":0.59,"cache_read":0.059},"canonical_model_id":"tencent/hy3"},"tencent/hy3-free":{"id":"tencent/hy3-free","name":"Hy3 (free)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"tencent/hy3"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-reasoner":{"id":"deepseek/deepseek-reasoner","name":"DeepSeek Reasoner","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.242,"output":0.726,"cache_read":0.0077},"canonical_model_id":"deepseek/deepseek-reasoner"},"deepseek/deepseek-v4-flash-free":{"id":"deepseek/deepseek-v4-flash-free","name":"DeepSeek V4 Flash (free)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"grok/grok-4.6":{"id":"grok/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"grok/grok-4.3":{"id":"grok/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"grok/grok-4.5":{"id":"grok/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.03,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5.2-chat-latest":{"id":"openai/gpt-5.2-chat-latest","name":"GPT-5.2 Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":100000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-5.1-chat-latest":{"id":"openai/gpt-5.1-chat-latest","name":"GPT-5.1 Chat","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5,"cache_read":0},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5-chat-latest":{"id":"openai/gpt-5-chat-latest","name":"GPT-5 Chat (latest)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":100000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-chat-latest"},"kimi/kimi-k2.7-code":{"id":"kimi/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi/kimi-k2.5":{"id":"kimi/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3,"cache_read":0.1,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"kimi/kimi-k3":{"id":"kimi/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"kimi/kimi-k2.6":{"id":"kimi/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":10},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.35,"output":1.42,"cache_read":0.071,"tiers":[{"input":1.06,"output":4.25,"cache_read":0.21,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.06,"output":4.25,"cache_read":0.21}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.276,"output":1.651,"cache_read":0.0276,"cache_write":0.345,"tiers":[{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.101,"output":6.602,"cache_read":0.1101,"cache_write":1.37625}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"tiers":[{"input":0.5,"output":3,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.5,"output":3}},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.172,"output":1.032,"tiers":[{"input":0.43,"output":2.58,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.115,"output":0.917,"tiers":[{"input":0.287,"output":2.294,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"tiers":[{"input":1,"output":4,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1,"output":4}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.359,"output":1.434,"tiers":[{"input":0.574,"output":2.294,"tier":{"type":"context","size":32000}},{"input":1.004,"output":4.014,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":3.75,"cache_read":0.25,"cache_write":1.563},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.057,"output":0.459,"tiers":[{"input":0.229,"output":1.835,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-vl-235b-a22b-thinking":{"id":"qwen/qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.4,"output":4},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.086,"output":0.688,"tiers":[{"input":0.258,"output":2.064,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.33,"output":2.4},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":131072},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"}}},"impossibl":{"id":"impossibl","env":["IMPOSSIBL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.impossibl.com/v1","name":"Impossibl","doc":"https://impossibl.com/docs/models","models":{"fireworks/gpt-oss-120b":{"id":"fireworks/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"},"fireworks/glm-5.2":{"id":"fireworks/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"fireworks/gpt-oss-20b":{"id":"fireworks/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.07,"output":0.3,"cache_read":0.035},"canonical_model_id":"openai/gpt-oss-20b"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.003},"canonical_model_id":"xiaomi/mimo-v2.5"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.19,"output":0.51,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"groq/gpt-oss-120b":{"id":"groq/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"groq/gpt-oss-20b":{"id":"groq/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"cache_read":3,"tiers":[{"input":60,"output":270,"cache_read":3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270,"cache_read":3}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"cache_read":3,"tiers":[{"input":60,"output":270,"cache_read":3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270,"cache_read":3}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"tier":{"type":"context","size":262144}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.8-max-preview":{"id":"qwen/qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":2.5,"output":7.5},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.05,"tiers":[{"input":1,"output":4,"cache_read":0.2,"tier":{"type":"context","size":262144}}],"context_over_200k":{"input":1,"output":4,"cache_read":0.2}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.7-max"}}},"edenai":{"id":"edenai","env":["EDENAI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.edenai.run/v3","name":"Eden AI","doc":"https://docs.edenai.co","models":{"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"google/deep-research-max-preview-04-2026":{"id":"google/deep-research-max-preview-04-2026","name":"Deep Research Max Preview","description":"Maximum-comprehensiveness agentic researcher for multi-step investigation, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/deep-research-max-preview-04-2026"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"reasoning":3,"cache_read":0.05,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-robotics-er-2-preview":{"id":"google/gemini-robotics-er-2-preview","name":"Gemini Robotics-ER 2 Preview","description":"Embodied reasoning model for robotics: spatial understanding, video progress, and multi-robot orchestration","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","pdf","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":1,"output":5,"reasoning":5,"cache_read":0.1,"input_audio":1},"canonical_model_id":"google/gemini-robotics-er-2-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"google/deep-research-preview-04-2026":{"id":"google/deep-research-preview-04-2026","name":"Gemini Deep Research Preview","description":"Agentic model for autonomous multi-step research, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/deep-research-preview-04-2026"},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["audio","image","text","video"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemini-pro-latest":{"id":"google/gemini-pro-latest","name":"Gemini Pro Latest (Gemini 4 Argon)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest (Gemini 3.8 Flash)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Image model for prompt-driven generation, multi-turn editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"cloudflare/@cf/mistralai/mistral-small-3.1-24b-instruct":{"id":"cloudflare/@cf/mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B (Cloudflare)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":102400},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"cloudflare/@cf/aisingapore/gemma-sea-lion-v4-27b-it":{"id":"cloudflare/@cf/aisingapore/gemma-sea-lion-v4-27b-it","name":"Gemma-SEA-LION-v4-27B-IT (Cloudflare)","description":"Gemma 3 27B tuned by AI Singapore for Southeast Asian languages and instruction following","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"aisingapore/gemma-sea-lion-v4-27b-it"},"cloudflare/@cf/meta/llama-3.2-1b-instruct":{"id":"cloudflare/@cf/meta/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct (Cloudflare)","description":"Compact instruction-tuned Llama model for multilingual dialogue and on-device assistants","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":32000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"cloudflare/@cf/meta/llama-guard-3-8b":{"id":"cloudflare/@cf/meta/llama-guard-3-8b","name":"Llama-Guard-3-8B (Cloudflare)","description":"Llama 3.1-based safety classifier for moderating prompts and model responses","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.484,"output":0.03},"canonical_model_id":"meta/llama-guard-3-8b"},"cloudflare/@cf/meta/llama-3.2-3b-instruct":{"id":"cloudflare/@cf/meta/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct (Cloudflare)","description":"Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":131072},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"cloudflare/@cf/deepseek-ai/deepseek-v4.1-flash-1017":{"id":"cloudflare/@cf/deepseek-ai/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Cloudflare)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"cloudflare/@cf/deepseek-ai/deepseek-v4-pro-0813":{"id":"cloudflare/@cf/deepseek-ai/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Cloudflare)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"cloudflare/@cf/deepseek-ai/deepseek-v4-flash-0731":{"id":"cloudflare/@cf/deepseek-ai/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Cloudflare)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"cloudflare/@cf/openai/gpt-oss-120b":{"id":"cloudflare/@cf/openai/gpt-oss-120b","name":"GPT OSS 120B (Cloudflare)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"cloudflare/@cf/openai/gpt-oss-20b":{"id":"cloudflare/@cf/openai/gpt-oss-20b","name":"GPT OSS 20B (Cloudflare)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.2,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"cloudflare/@cf/qwen/qwen2.5-coder-32b-instruct":{"id":"cloudflare/@cf/qwen/qwen2.5-coder-32b-instruct","name":"Qwen2.5-Coder-32B-Instruct (Cloudflare)","description":"Open coding-focused Qwen model for code generation, repair, and repository reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"cloudflare/@cf/zai-org/glm-4.7-flash":{"id":"cloudflare/@cf/zai-org/glm-4.7-flash","name":"GLM-4.7-Flash (Cloudflare)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"ovhcloud/gpt-oss-120b":{"id":"ovhcloud/gpt-oss-120b","name":"GPT OSS 120B (OVHcloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.09,"output":0.47},"canonical_model_id":"openai/gpt-oss-120b"},"ovhcloud/gpt-oss-20b":{"id":"ovhcloud/gpt-oss-20b","name":"GPT OSS 20B (OVHcloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"greenference/deepseek-v4.1-flash":{"id":"greenference/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Greenference)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.22,"output":0.88,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"greenference/gpt-oss-20b":{"id":"greenference/gpt-oss-20b","name":"GPT OSS 20B (Greenference)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.009,"output":0.045,"cache_read":0.0045},"canonical_model_id":"openai/gpt-oss-20b"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4-fast-non-reasoning":{"id":"xai/grok-4-fast-non-reasoning","name":"X-Ai/Grok-4-Fast-Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05,"tiers":[{"input":0.4,"output":1,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-latest":{"id":"xai/grok-latest","name":"Grok Latest (Grok 4.7)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"tencent/deepseek-v4.1-flash":{"id":"tencent/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Tencent)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview (Tencent)","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/step-5-preview":{"id":"tencent/step-5-preview","name":"Step 5 Preview (Tencent)","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"tencent/deepseek-v4-pro-0813":{"id":"tencent/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Tencent)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3 (Tencent)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"greenpt/gpt-oss-120b":{"id":"greenpt/gpt-oss-120b","name":"GPT OSS 120B (Greenpt)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.22386,"output":0.78351},"canonical_model_id":"openai/gpt-oss-120b"},"greenpt/llama-3.3-70b-instruct":{"id":"greenpt/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct (Greenpt)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":1.231229,"output":1.231229},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"greenpt/deepseek-v4.1-flash":{"id":"greenpt/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Greenpt)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.246246,"output":1.231229,"cache_read":0.012312},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"greenpt/qwen3-235b-a22b-instruct-2507":{"id":"greenpt/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Greenpt)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":250000,"output":16384},"cost":{"input":1.007369,"output":3.022108},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"greenpt/deepseek-v4-flash-0731":{"id":"greenpt/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Greenpt)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.156702,"output":0.391755,"cache_read":0.044772},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder (Meta)","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1 (Meta)","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3 (Meta)","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2 (Meta)","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"together_ai/thinkingmachines/Inkling":{"id":"together_ai/thinkingmachines/Inkling","name":"Inkling (Together AI)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"together_ai/reflection/Beam-1.0-501B-A23B":{"id":"together_ai/reflection/Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B (Together AI)","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":1.2,"cache_read":0.04},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"together_ai/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"together_ai/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Together AI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"together_ai/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"together_ai/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Together AI)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"together_ai/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"together_ai/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Together AI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"together_ai/deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"together_ai/deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813 (Together AI)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"together_ai/openai/gpt-oss-120b":{"id":"together_ai/openai/gpt-oss-120b","name":"GPT OSS 120B (Together AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"together_ai/meta-models/Muse-Glimmer-30B":{"id":"together_ai/meta-models/Muse-Glimmer-30B","name":"Muse Glimmer 30B (Together AI)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"fireworks_ai/gpt-oss-120b":{"id":"fireworks_ai/gpt-oss-120b","name":"GPT OSS 120B (Fireworks AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.014},"canonical_model_id":"openai/gpt-oss-120b"},"fireworks_ai/accounts/fireworks/models/gpt-oss-120b":{"id":"fireworks_ai/accounts/fireworks/models/gpt-oss-120b","name":"GPT OSS 120B (Fireworks AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"},"fireworks_ai/accounts/fireworks/models/ember-1":{"id":"fireworks_ai/accounts/fireworks/models/ember-1","name":"Ember-1 (Fireworks AI)","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"fireworks_ai/accounts/fireworks/models/inkling":{"id":"fireworks_ai/accounts/fireworks/models/inkling","name":"Inkling (Fireworks AI)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"fireworks_ai/accounts/fireworks/models/muse-spark-open":{"id":"fireworks_ai/accounts/fireworks/models/muse-spark-open","name":"Muse Spark Open (Fireworks AI)","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"scaleway/gemma-3-27b-it":{"id":"scaleway/gemma-3-27b-it","name":"Gemma 3 27B IT (Scaleway)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":131072},"cost":{"input":0.287125,"output":0.57425},"canonical_model_id":"google/gemma-3-27b-it"},"scaleway/gpt-oss-120b":{"id":"scaleway/gpt-oss-120b","name":"GPT OSS 120B (Scaleway)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.167895,"output":0.67158},"canonical_model_id":"openai/gpt-oss-120b"},"scaleway/llama-3.3-70b-instruct":{"id":"scaleway/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct (Scaleway)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":1.007369,"output":1.007369},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"scaleway/qwen3-coder-30b-a3b-instruct":{"id":"scaleway/qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct (Scaleway)","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":65536},"cost":{"input":0.2245,"output":0.898},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"scaleway/deepseek-v4.1-flash":{"id":"scaleway/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Scaleway)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":384000},"cost":{"input":0.67158,"output":2.014739,"cache_read":0.134316},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"scaleway/qwen3-235b-a22b-instruct-2507":{"id":"scaleway/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Scaleway)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":250000,"output":16384},"cost":{"input":0.839475,"output":2.518424},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"scaleway/deepseek-v4-flash-0731":{"id":"scaleway/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Scaleway)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":384000},"cost":{"input":0.44772,"output":0.89544,"cache_read":0.089544},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"tensorx/deepseek/deepseek-v4.1-flash":{"id":"tensorx/deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (TensorX)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.125},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"tensorx/deepseek/deepseek-v4.1-flash-1017":{"id":"tensorx/deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (TensorX)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.125},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"tensorx/deepseek/deepseek-v4-pro-0813":{"id":"tensorx/deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (TensorX)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":2,"output":4,"cache_read":0.5},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"tensorx/deepseek/deepseek-v4-flash-0731":{"id":"tensorx/deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (TensorX)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.25,"output":0.3,"cache_read":0.0625},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"tensorx/deepseek/deepseek-r1-0528":{"id":"tensorx/deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528 (TensorX)","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":32768},"cost":{"input":0.66,"output":2.6,"cache_read":0.165},"canonical_model_id":"deepseek/deepseek-r1-0528"},"tensorx/moonshotai/kimi-k2.5":{"id":"tensorx/moonshotai/kimi-k2.5","name":"Kimi K2.5 (TensorX)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.8,"cache_read":0.125},"canonical_model_id":"moonshotai/kimi-k2.5"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B (Cerebras)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.35,"output":0.75,"cache_read":0.35},"canonical_model_id":"openai/gpt-oss-120b"},"nebius/google/gemma-3-27b-it":{"id":"nebius/google/gemma-3-27b-it","name":"Gemma 3 27B IT (Nebius)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":110000,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"google/gemma-3-27b-it"},"nebius/nvidia/Nemotron-3-Ultra-550b-a55b":{"id":"nebius/nvidia/Nemotron-3-Ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B (Nebius)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":1,"output":3,"cache_read":1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nebius/nvidia/nemotron-3-super-120b-a12b":{"id":"nebius/nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B (Nebius)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":0.9,"cache_read":0.3},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nebius/Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"nebius/Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Nebius)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.2,"output":0.6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"nebius/Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"nebius/Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3 30B-A3B Instruct 2507 (Nebius)","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"nebius/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"nebius/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Nebius)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"nebius/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"nebius/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Nebius)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"nebius/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"nebius/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Nebius)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"nebius/deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"nebius/deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813 (Nebius)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":979000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":1.32},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"nebius/openai/gpt-oss-120b":{"id":"nebius/openai/gpt-oss-120b","name":"GPT OSS 120B (Nebius)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.15},"canonical_model_id":"openai/gpt-oss-120b"},"nebius/meta-models/muse-spark-open":{"id":"nebius/meta-models/muse-spark-open","name":"Muse Spark Open (Nebius)","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1.25},"canonical_model_id":"meta/muse-spark-open"},"vertex/gemini-3.1-flash-image":{"id":"vertex/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2 (Vertex AI)","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"vertex/gemini-3.5-flash-lite@eu":{"id":"vertex/gemini-3.5-flash-lite@eu","name":"Gemini 3.5 Flash Lite (Vertex AI, EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"vertex/gemini-4-argon":{"id":"vertex/gemini-4-argon","name":"Gemini 4 Argon (Vertex AI)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-3-flash-preview":{"id":"vertex/gemini-3-flash-preview","name":"Gemini 3 Flash Preview (Vertex AI)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"reasoning":3,"cache_read":0.05,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"vertex/gemini-3.5-flash":{"id":"vertex/gemini-3.5-flash","name":"Gemini 3.5 Flash (Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"vertex/gemini-3.8-flash":{"id":"vertex/gemini-3.8-flash","name":"Gemini 3.8 Flash (Vertex AI)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-3.1-pro-preview":{"id":"vertex/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview (Vertex AI)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":0.25}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"vertex/gemini-3.7-flash":{"id":"vertex/gemini-3.7-flash","name":"Gemini 3.7 Flash (Vertex AI)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"vertex/gemini-3.7-flash@us":{"id":"vertex/gemini-3.7-flash@us","name":"Gemini 3.7 Flash (Vertex AI, US)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"vertex/gemini-3.1-flash-lite-image":{"id":"vertex/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite (Vertex AI)","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"vertex/gemini-2.5-flash-image":{"id":"vertex/gemini-2.5-flash-image","type":"image","name":"Nano Banana (Vertex AI)","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash-image"},"vertex/gemini-3.5-flash@us":{"id":"vertex/gemini-3.5-flash@us","name":"Gemini 3.5 Flash (Vertex AI, US)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"vertex/gemini-3.6-flash@eu":{"id":"vertex/gemini-3.6-flash@eu","name":"Gemini 3.6 Flash (Vertex AI, EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"vertex/gemini-3.5-flash@eu":{"id":"vertex/gemini-3.5-flash@eu","name":"Gemini 3.5 Flash (Vertex AI, EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"vertex/gemini-4-argon@us":{"id":"vertex/gemini-4-argon@us","name":"Gemini 4 Argon (Vertex AI, US)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-3.5-flash-lite@us":{"id":"vertex/gemini-3.5-flash-lite@us","name":"Gemini 3.5 Flash Lite (Vertex AI, US)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"vertex/gemini-3-pro-image":{"id":"vertex/gemini-3-pro-image","type":"image","name":"Nano Banana Pro (Vertex AI)","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image"},"vertex/gemini-pro-latest":{"id":"vertex/gemini-pro-latest","name":"Gemini Pro Latest (Gemini 4 Argon, Vertex AI)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-3.7-flash@eu":{"id":"vertex/gemini-3.7-flash@eu","name":"Gemini 3.7 Flash (Vertex AI, EU)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"vertex/gemini-3.8-flash@us":{"id":"vertex/gemini-3.8-flash@us","name":"Gemini 3.8 Flash (Vertex AI, US)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-3.6-flash":{"id":"vertex/gemini-3.6-flash","name":"Gemini 3.6 Flash (Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"vertex/gemini-3.8-flash@eu":{"id":"vertex/gemini-3.8-flash@eu","name":"Gemini 3.8 Flash (Vertex AI, EU)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-4-argon@eu":{"id":"vertex/gemini-4-argon@eu","name":"Gemini 4 Argon (Vertex AI, EU)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":20,"reasoning":20,"cache_read":0.2,"cache_write":0.375,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"},"vertex/gemini-flash-latest":{"id":"vertex/gemini-flash-latest","name":"Gemini Flash Latest (Gemini 3.8 Flash, Vertex AI)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"reasoning":7.5,"cache_read":0.15,"cache_write":0.083333,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"vertex/gemini-3.5-flash-lite":{"id":"vertex/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite (Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"vertex/gemini-3.1-flash-lite@eu":{"id":"vertex/gemini-3.1-flash-lite@eu","name":"Gemini 3.1 Flash Lite (Vertex AI, EU)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"vertex/gemini-3.6-flash@us":{"id":"vertex/gemini-3.6-flash@us","name":"Gemini 3.6 Flash (Vertex AI, US)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"vertex/gemini-3.1-flash-lite@us":{"id":"vertex/gemini-3.1-flash-lite@us","name":"Gemini 3.1 Flash Lite (Vertex AI, US)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"vertex/gemini-nano-banana-2.1":{"id":"vertex/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1 (Vertex AI)","description":"Image model for prompt-driven generation, multi-turn editing, and visual design workflows","family":"gemini-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"vertex/gemini-3.1-flash-lite":{"id":"vertex/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite (Vertex AI)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"groq/meta-llama/llama-prompt-guard-2-86m":{"id":"groq/meta-llama/llama-prompt-guard-2-86m","name":"Llama Prompt Guard 2 86M (Groq)","description":"86M-parameter (mDeBERTa-base) multilingual classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"cost":{"input":0.04,"output":0.04,"cache_read":0.02},"canonical_model_id":"meta/llama-prompt-guard-2-86m"},"groq/meta-llama/llama-prompt-guard-2-22m":{"id":"groq/meta-llama/llama-prompt-guard-2-22m","name":"Llama Prompt Guard 2 22M (Groq)","description":"22M-parameter (DeBERTa-xsmall) English classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"cost":{"input":0.03,"output":0.03,"cache_read":0.015},"canonical_model_id":"meta/llama-prompt-guard-2-22m"},"groq/openai/gpt-oss-120b":{"id":"groq/openai/gpt-oss-120b","name":"GPT OSS 120B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"groq/openai/gpt-oss-20b":{"id":"groq/openai/gpt-oss-20b","name":"GPT OSS 20B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"groq/openai/gpt-oss-safeguard-20b":{"id":"groq/openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B (Groq)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"amazon/zai.glm-4.7-flash":{"id":"amazon/zai.glm-4.7-flash","name":"GLM-4.7-Flash (Amazon Bedrock)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"amazon/openai.gpt-oss-safeguard-20b":{"id":"amazon/openai.gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B (Amazon Bedrock)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"amazon/mistral.voxtral-small-24b-2507":{"id":"amazon/mistral.voxtral-small-24b-2507","name":"Voxtral Small 24B 2507 (Amazon Bedrock)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"amazon/amazon.nova-micro-v1:0":{"id":"amazon/amazon.nova-micro-v1:0","name":"Nova Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875},"canonical_model_id":"amazon/nova-micro"},"amazon/amazon.nova-lite-v1:0":{"id":"amazon/amazon.nova-lite-v1:0","name":"Nova Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015},"canonical_model_id":"amazon/nova-lite"},"amazon/google.gemma-3-12b-it":{"id":"amazon/google.gemma-3-12b-it","name":"Gemma 3 12B IT (Amazon Bedrock)","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"amazon/mistral.voxtral-mini-3b-2507":{"id":"amazon/mistral.voxtral-mini-3b-2507","name":"Voxtral Mini 3B 2507 (Amazon Bedrock)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.04,"output":0.04},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"amazon/moonshotai.kimi-k2.5":{"id":"amazon/moonshotai.kimi-k2.5","name":"Kimi K2.5 (Amazon Bedrock)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"amazon/google.gemma-3-27b-it":{"id":"amazon/google.gemma-3-27b-it","name":"Gemma 3 27B IT (Amazon Bedrock)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.23,"output":0.38},"canonical_model_id":"google/gemma-3-27b-it"},"amazon/moonshot.kimi-k2-thinking":{"id":"amazon/moonshot.kimi-k2-thinking","name":"Kimi K2 Thinking (Amazon Bedrock)","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"amazon/google.gemma-3-12b-it@us":{"id":"amazon/google.gemma-3-12b-it@us","name":"Gemma 3 12B IT (Amazon Bedrock, US)","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"amazon/qwen.qwen3-coder-next":{"id":"amazon/qwen.qwen3-coder-next","name":"Qwen3 Coder Next (Amazon Bedrock)","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"amazon/mistral.voxtral-small-24b-2507@us":{"id":"amazon/mistral.voxtral-small-24b-2507@us","name":"Voxtral Small 24B 2507 (Amazon Bedrock, US)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"amazon/google.gemma-3-4b-it@us":{"id":"amazon/google.gemma-3-4b-it@us","name":"Gemma 3 4B IT (Amazon Bedrock, US)","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"amazon/google.gemma-3-27b-it@us":{"id":"amazon/google.gemma-3-27b-it@us","name":"Gemma 3 27B IT (Amazon Bedrock, US)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.23,"output":0.38},"canonical_model_id":"google/gemma-3-27b-it"},"amazon/amazon.nova-pro-v1:0":{"id":"amazon/amazon.nova-pro-v1:0","name":"Nova Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2},"canonical_model_id":"amazon/nova-pro"},"amazon/google.gemma-3-4b-it":{"id":"amazon/google.gemma-3-4b-it","name":"Gemma 3 4B IT (Amazon Bedrock)","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"amazon/zai.glm-4.7-flash@us":{"id":"amazon/zai.glm-4.7-flash@us","name":"GLM-4.7-Flash (Amazon Bedrock, US)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"amazon/amazon.nova-lite-v1:0@us":{"id":"amazon/amazon.nova-lite-v1:0@us","name":"Nova Lite (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015},"canonical_model_id":"amazon/nova-lite"},"amazon/mistral.voxtral-mini-3b-2507@us":{"id":"amazon/mistral.voxtral-mini-3b-2507@us","name":"Voxtral Mini 3B 2507 (Amazon Bedrock, US)","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.04,"output":0.04},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"amazon/mistral.pixtral-large-2502-v1:0":{"id":"amazon/mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02) (Amazon Bedrock)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"amazon/amazon.nova-pro-v1:0@us":{"id":"amazon/amazon.nova-pro-v1:0@us","name":"Nova Pro (US)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2},"canonical_model_id":"amazon/nova-pro"},"amazon/openai.gpt-oss-safeguard-20b@us":{"id":"amazon/openai.gpt-oss-safeguard-20b@us","name":"GPT OSS Safeguard 20B (Amazon Bedrock, US)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":131072},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"amazon/mistral.pixtral-large-2502-v1:0@us":{"id":"amazon/mistral.pixtral-large-2502-v1:0@us","name":"Pixtral Large (25.02) (Amazon Bedrock, US)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"amazon/amazon.nova-micro-v1:0@us":{"id":"amazon/amazon.nova-micro-v1:0@us","name":"Nova Micro (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875},"canonical_model_id":"amazon/nova-micro"},"ionos/meta-llama/Llama-3.3-70B-Instruct":{"id":"ionos/meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct (IONOS)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.727545,"output":0.727545},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"ionos/Qwen/Qwen3-Coder-Next":{"id":"ionos/Qwen/Qwen3-Coder-Next","name":"Qwen3 Coder Next (IONOS)","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":65536},"cost":{"input":0.167895,"output":0.89544},"canonical_model_id":"alibaba/qwen3-coder-next"},"ionos/openai/gpt-oss-120b":{"id":"ionos/openai/gpt-oss-120b","name":"GPT OSS 120B (IONOS)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.167895,"output":0.727545},"canonical_model_id":"openai/gpt-oss-120b"},"infomaniak/mistralai/Ministral-3-14B-Instruct-2512":{"id":"infomaniak/mistralai/Ministral-3-14B-Instruct-2512","name":"Ministral 3 14B (Infomaniak)","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":262144},"cost":{"input":0.33579,"output":0.44772},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"databricks/databricks-qwen3-next-80b-a3b-instruct":{"id":"databricks/databricks-qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct (Databricks)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15001,"output":1.20001,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"databricks/databricks-deepseek-v4-flash-0731":{"id":"databricks/databricks-deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Databricks)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.014,"cache_write":0.14},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"databricks/databricks-qwen3-next-80b-a3b-instruct@eu":{"id":"databricks/databricks-qwen3-next-80b-a3b-instruct@eu","name":"Qwen3-Next 80B-A3B Instruct (Databricks, EU)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15001,"output":1.20001,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"databricks/databricks-deepseek-v4-pro-0813":{"id":"databricks/databricks-deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Databricks)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.959999,"cache_read":0.132,"cache_write":1.31999},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"databricks/databricks-inkling":{"id":"databricks/databricks-inkling","name":"Inkling (Databricks)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1048576},"cost":{"input":1,"output":4.05,"cache_read":0.1,"cache_write":1.00002},"canonical_model_id":"thinkingmachines/inkling"},"databricks/databricks-gpt-oss-20b":{"id":"databricks/databricks-gpt-oss-20b","name":"GPT OSS 20B (Databricks)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.30002,"cache_read":0.007,"cache_write":0.07},"canonical_model_id":"openai/gpt-oss-20b"},"databricks/databricks-gpt-oss-20b@eu":{"id":"databricks/databricks-gpt-oss-20b@eu","name":"GPT OSS 20B (Databricks, EU)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.30002,"cache_read":0.007,"cache_write":0.07},"canonical_model_id":"openai/gpt-oss-20b"},"databricks/databricks-gpt-oss-120b@eu":{"id":"databricks/databricks-gpt-oss-120b@eu","name":"GPT OSS 120B (Databricks, EU)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15001,"output":0.59997,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"openai/gpt-oss-120b"},"databricks/databricks-gpt-oss-120b":{"id":"databricks/databricks-gpt-oss-120b","name":"GPT OSS 120B (Databricks)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15001,"output":0.59997,"cache_read":0.015001,"cache_write":0.15001},"canonical_model_id":"openai/gpt-oss-120b"},"perplexityai/sonar":{"id":"perplexityai/sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127072,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"perplexityai/sonar-reasoning-pro":{"id":"perplexityai/sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"perplexityai/sonar-deep-research":{"id":"perplexityai/sonar-deep-research","name":"Sonar Deep Research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"},"perplexityai/sonar-pro":{"id":"perplexityai/sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":150,"output":600},"canonical_model_id":"openai/o1-pro"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-latest":{"id":"openai/gpt-latest","name":"GPT Latest (GPT-6 Astra)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-5-search-api":{"id":"openai/gpt-5-search-api","type":"chat","name":"GPT-5 Search API","description":"GPT-5 model trained for web search in the Chat Completions API","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":100000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-search-api"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-pro-latest":{"id":"openai/gpt-pro-latest","name":"GPT Pro Latest (GPT-5.6 Sol)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":8192},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-mini-latest":{"id":"openai/gpt-mini-latest","name":"GPT Mini Latest (GPT-5.6 Luna)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","pdf","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"zai/glm-5v-turbo":{"id":"zai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-4.6v":{"id":"zai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"sference/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"sference/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Sference)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"sference/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"sference/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Sference)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"azure/gpt-5.1-codex-max":{"id":"azure/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"azure/gpt-5.1-codex":{"id":"azure/gpt-5.1-codex","name":"GPT-5.1 Codex (Azure)","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"azure/gpt-5.1-codex-mini":{"id":"azure/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"azure/gpt-5.2-codex":{"id":"azure/gpt-5.2-codex","name":"GPT-5.2 Codex (Azure)","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-latest":{"id":"anthropic/claude-fable-latest","name":"Claude Fable Latest (Claude Fable 5.1)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-latest":{"id":"anthropic/claude-sonnet-latest","name":"Claude Sonnet Latest (Claude Sonnet 5.5)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-haiku-latest":{"id":"anthropic/claude-haiku-latest","name":"Claude Haiku Latest (Claude Haiku 5.5)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-latest":{"id":"anthropic/claude-opus-latest","name":"Claude Opus Latest (Claude Opus 5.5)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"minimax/MiniMax-M3.1-Flash":{"id":"minimax/MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/MiniMax-M2.1":{"id":"minimax/MiniMax-M2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/MiniMax-M2":{"id":"minimax/MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"minimax/MiniMax-M3":{"id":"minimax/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/MiniMax-M2.7":{"id":"minimax/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/MiniMax-M2.5":{"id":"minimax/MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"mistral/mistral-large-2407":{"id":"mistral/mistral-large-2407","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-2407"},"mistral/ministral-14b-2512":{"id":"mistral/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b-2512"},"mistral/devstral-medium-latest":{"id":"mistral/devstral-medium-latest","name":"Devstral 2 (latest)","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-medium-latest"},"mistral/devstral-2512":{"id":"mistral/devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistral/magistral-medium-latest":{"id":"mistral/magistral-medium-latest","name":"Magistral Medium (latest)","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-03-17","last_updated":"2025-03-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":16384},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/magistral-medium-latest"},"mistral/codestral-2508":{"id":"mistral/codestral-2508","name":"Codestral 25.08","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"mistral/ministral-8b-2512":{"id":"mistral/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-8b-2512"},"mistral/mistral-small-latest":{"id":"mistral/mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-latest"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/voxtral-small-latest":{"id":"mistral/voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"mistral/voxtral-small-latest"},"mistral/mistral-medium-latest":{"id":"mistral/mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-latest"},"mistral/mistral-small-2603":{"id":"mistral/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistral/codestral-latest":{"id":"mistral/codestral-latest","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-latest"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/mistral-medium-2604":{"id":"mistral/mistral-medium-2604","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-2604"},"mistral/mistral-medium-3-5":{"id":"mistral/mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistral/mistral-medium-3":{"id":"mistral/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3"},"mistral/mistral-medium-2505":{"id":"mistral/mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-2505"},"mistral/ministral-3b-2512":{"id":"mistral/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3b-2512"},"deepinfra/nemotron-3-super-120b-a12b":{"id":"deepinfra/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B (Deep Infra)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.085,"output":0.4},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"deepinfra/nemotron-3-ultra-550b-a55b":{"id":"deepinfra/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B (Deep Infra)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.5,"output":2.2,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"deepinfra/Gryphe/MythoMax-L2-13b":{"id":"deepinfra/Gryphe/MythoMax-L2-13b","name":"MythoMax L2 13B (Deep Infra)","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":3200},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"gryphe/mythomax-l2-13b"},"deepinfra/mistralai/Mixtral-8x7B-Instruct-v0.1":{"id":"deepinfra/mistralai/Mixtral-8x7B-Instruct-v0.1","name":"Mixtral 8x7B Instruct v0.1 (Deep Infra)","description":"Open sparse mixture-of-experts Mistral instruction model","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"mistral/mixtral-8x7b-instruct-v0.1"},"deepinfra/mistralai/Mistral-Small-24B-Instruct-2501":{"id":"deepinfra/mistralai/Mistral-Small-24B-Instruct-2501","name":"Mistral Small 3 (Deep Infra)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"deepinfra/google/gemma-3-27b-it":{"id":"deepinfra/google/gemma-3-27b-it","name":"Gemma 3 27B IT (Deep Infra)","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"deepinfra/google/gemma-3-12b-it":{"id":"deepinfra/google/gemma-3-12b-it","name":"Gemma 3 12B IT (Deep Infra)","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"deepinfra/google/gemma-3-4b-it":{"id":"deepinfra/google/gemma-3-4b-it","name":"Gemma 3 4B IT (Deep Infra)","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-4b-it"},"deepinfra/nvidia/Nemotron-3-Nano-30B-A3B":{"id":"deepinfra/nvidia/Nemotron-3-Nano-30B-A3B","name":"Nemotron 3 Nano 30B A3B (Deep Infra)","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"deepinfra/nvidia/Llama-3.1-Nemotron-70B-Instruct":{"id":"deepinfra/nvidia/Llama-3.1-Nemotron-70B-Instruct","name":"Llama 3.1 Nemotron 70B Instruct (Deep Infra)","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.6,"output":0.6},"canonical_model_id":"nvidia/llama-3.1-nemotron-70b-instruct"},"deepinfra/thinkingmachines/Inkling-Small":{"id":"deepinfra/thinkingmachines/Inkling-Small","name":"Inkling Small (Deep Infra)","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"deepinfra/thinkingmachines/Inkling":{"id":"deepinfra/thinkingmachines/Inkling","name":"Inkling (Deep Infra)","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.95,"output":4.05,"cache_read":0.16},"canonical_model_id":"thinkingmachines/inkling"},"deepinfra/tencent/Hy4-preview":{"id":"deepinfra/tencent/Hy4-preview","name":"Hy4 preview (Deep Infra)","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"deepinfra/tencent/Hy3":{"id":"deepinfra/tencent/Hy3","name":"Hy3 (Deep Infra)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.13,"output":0.53,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"deepinfra/meta-llama/Llama-3.2-11B-Vision-Instruct":{"id":"deepinfra/meta-llama/Llama-3.2-11B-Vision-Instruct","name":"Llama-3.2-11B-Vision-Instruct (Deep Infra)","description":"Open multimodal Llama model for image understanding, captioning, and visual QA","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.345,"output":0.345},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"deepinfra/meta-llama/Llama-Guard-4-12B":{"id":"deepinfra/meta-llama/Llama-Guard-4-12B","name":"Llama Guard 4 12B (Deep Infra)","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"deepinfra/meta-llama/Llama-3.2-3B-Instruct":{"id":"deepinfra/meta-llama/Llama-3.2-3B-Instruct","name":"Llama 3.2 3B Instruct (Deep Infra)","description":"Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.02,"output":0.02},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"deepinfra/meta-llama/Llama-Guard-3-8B":{"id":"deepinfra/meta-llama/Llama-Guard-3-8B","name":"Llama-Guard-3-8B (Deep Infra)","description":"Llama 3.1-based safety classifier for moderating prompts and model responses","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.055,"output":0.055},"canonical_model_id":"meta/llama-guard-3-8b"},"deepinfra/meta-llama/Llama-3.3-70B-Instruct":{"id":"deepinfra/meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct (Deep Infra)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"deepinfra/moonshotai/Kimi-K2-Instruct":{"id":"deepinfra/moonshotai/Kimi-K2-Instruct","name":"Kimi K2 Instruct (Deep Infra)","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"deepinfra/moonshotai/Kimi-K2.5":{"id":"deepinfra/moonshotai/Kimi-K2.5","name":"Kimi K2.5 (Deep Infra)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"deepinfra/Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"deepinfra/Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3 VL 30B A3B Instruct (Deep Infra)","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen3-Coder 480B-A35B Instruct (Deep Infra)","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"deepinfra/Qwen/Qwen3-VL-235B-A22B-Instruct":{"id":"deepinfra/Qwen/Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct (Deep Infra)","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"deepinfra/Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"deepinfra/Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Deep Infra)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3 235B-A22B Thinking 2507 (Deep Infra)","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.23,"output":2.3,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"deepinfra/Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"deepinfra/Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next 80B-A3B Instruct (Deep Infra)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"deepinfra/Qwen/Qwen2.5-7B-Instruct":{"id":"deepinfra/Qwen/Qwen2.5-7B-Instruct","name":"Qwen2.5 7B Instruct (Deep Infra)","description":"Small open Qwen2.5 instruction model for chat, structured output, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.04,"output":0.1},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"deepinfra/Qwen/Qwen3-Max":{"id":"deepinfra/Qwen/Qwen3-Max","name":"Qwen3 Max (Deep Infra)","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1.2,"output":6,"cache_read":0.24},"canonical_model_id":"alibaba/qwen3-max"},"deepinfra/ibm-granite/granite-4.2-8b":{"id":"deepinfra/ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B (Deep Infra)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"deepinfra/deepseek-ai/DeepSeek-V3-0324":{"id":"deepinfra/deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324 (Deep Infra)","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.24,"output":0.9,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Deep Infra)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepinfra/deepseek-ai/DeepSeek-R1":{"id":"deepinfra/deepseek-ai/DeepSeek-R1","name":"DeepSeek-R1 (Deep Infra)","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.7,"output":2.4},"canonical_model_id":"deepseek/deepseek-r1"},"deepinfra/deepseek-ai/DeepSeek-V3":{"id":"deepinfra/deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3 (Deep Infra)","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":8192},"cost":{"input":0.32,"output":0.89},"canonical_model_id":"deepseek/deepseek-v3"},"deepinfra/deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepinfra/deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (Deep Infra)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.06,"output":0.18,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepinfra/deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (Deep Infra)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepinfra/deepseek-ai/DeepSeek-R1-0528":{"id":"deepinfra/deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528 (Deep Infra)","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepinfra/deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepinfra/deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813 (Deep Infra)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepinfra/openai/gpt-oss-120b":{"id":"deepinfra/openai/gpt-oss-120b","name":"GPT OSS 120B (Deep Infra)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.037,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"deepinfra/openai/gpt-oss-20b":{"id":"deepinfra/openai/gpt-oss-20b","name":"GPT OSS 20B (Deep Infra)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.03,"output":0.14},"canonical_model_id":"openai/gpt-oss-20b"},"deepinfra/stepfun-ai/Step-3.7-Flash":{"id":"deepinfra/stepfun-ai/Step-3.7-Flash","name":"Step 3.7 Flash (Deep Infra)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"deepinfra/stepfun-ai/Step-5-Preview":{"id":"deepinfra/stepfun-ai/Step-5-Preview","name":"Step 5 Preview (Deep Infra)","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"deepinfra/stepfun-ai/Step-3.5-Flash":{"id":"deepinfra/stepfun-ai/Step-3.5-Flash","name":"Step 3.5 Flash (Deep Infra)","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.09,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"deepinfra/microsoft/WizardLM-2-8x22B":{"id":"deepinfra/microsoft/WizardLM-2-8x22B","name":"WizardLM 2 8x22B (Deep Infra)","description":"Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-04-15","last_updated":"2024-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8000},"cost":{"input":0.48,"output":0.48},"canonical_model_id":"microsoft/wizardlm-2-8x22b"},"deepinfra/zai-org/GLM-4.7-Flash":{"id":"deepinfra/zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash (Deep Infra)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"deepinfra/meta-models/Muse-Glimmer-30B":{"id":"deepinfra/meta-models/Muse-Glimmer-30B","name":"Muse Glimmer 30B (Deep Infra)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"deepinfra/ByteDance/Seed-2.0-code":{"id":"deepinfra/ByteDance/Seed-2.0-code","name":"Seed 2.0 Code (Deep Infra)","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"deepinfra/ByteDance/Seed-2.0-mini":{"id":"deepinfra/ByteDance/Seed-2.0-mini","name":"Seed 2.0 Mini (Deep Infra)","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"deepinfra/NousResearch/Hermes-3-Llama-3.1-70B":{"id":"deepinfra/NousResearch/Hermes-3-Llama-3.1-70B","name":"Hermes 3 70B Instruct (Deep Infra)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.7,"output":0.7},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"deepinfra/NousResearch/Hermes-3-Llama-3.1-405B":{"id":"deepinfra/NousResearch/Hermes-3-Llama-3.1-405B","name":"Hermes 3 405B Instruct (Deep Infra)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":1,"output":1},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"qwen/deepseek-v4.1-flash":{"id":"qwen/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Alibaba)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"qwen/deepseek-v4.1-flash-1017":{"id":"qwen/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Alibaba)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"qwen/qwen-mt-plus":{"id":"qwen/qwen-mt-plus","name":"Qwen MT Plus","description":"Flagship Qwen translation model with terminology control and support for 92 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-24","last_updated":"2025-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":8192},"cost":{"input":2.46,"output":7.37},"canonical_model_id":"alibaba/qwen-mt-plus"},"qwen/qwen-max":{"id":"qwen/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4,"cache_read":0.32},"canonical_model_id":"alibaba/qwen-max"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/deepseek-v4-pro-0813":{"id":"qwen/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813 (Alibaba)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.066},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"qwen/deepseek-v4-flash-0731":{"id":"qwen/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (Alibaba)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.8-27b"},"arkanecloud/ministral-14b":{"id":"arkanecloud/ministral-14b","name":"Ministral 14B (Arkanecloud)","description":"Compact multimodal Mistral model for local assistants, edge agents, and efficient tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":262144},"cost":{"input":0.15,"output":0.15,"cache_read":0.15},"canonical_model_id":"mistral/ministral-14b"},"arkanecloud/nvidia/nemotron-3-super-120b-a12b":{"id":"arkanecloud/nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B (Arkanecloud)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"cost":{"input":0.4,"output":0.8,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"arkanecloud/nvidia/nemotron-3-ultra-550b-a55b":{"id":"arkanecloud/nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B (Arkanecloud)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512288,"output":128000},"cost":{"input":0.6,"output":3.5,"cache_read":0.2},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"arkanecloud/meta-llama/Llama-3.1-70B-Instruct":{"id":"arkanecloud/meta-llama/Llama-3.1-70B-Instruct","name":"Llama-3.1-70B-Instruct (Arkanecloud)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.8,"output":0.8,"cache_read":0.16},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"arkanecloud/meta-llama/Llama-3.3-70B-Instruct":{"id":"arkanecloud/meta-llama/Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct (Arkanecloud)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.7,"output":0.7,"cache_read":0.2},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"arkanecloud/meta-llama/Llama-3.1-8B-Instruct":{"id":"arkanecloud/meta-llama/Llama-3.1-8B-Instruct","name":"Llama-3.1-8B-Instruct (Arkanecloud)","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.15,"output":0.15,"cache_read":0.03},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"arkanecloud/Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"arkanecloud/Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507 (Arkanecloud)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":16384},"cost":{"input":0.25,"output":1,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"arkanecloud/deepseek-ai/DeepSeek-V4.1-Flash":{"id":"arkanecloud/deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (Arkanecloud)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.075},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"arkanecloud/openai/gpt-oss-120b":{"id":"arkanecloud/openai/gpt-oss-120b","name":"GPT OSS 120B (Arkanecloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-oss-120b"},"arkanecloud/openai/gpt-oss-20b":{"id":"arkanecloud/openai/gpt-oss-20b","name":"GPT OSS 20B (Arkanecloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.25,"cache_read":0.05},"canonical_model_id":"openai/gpt-oss-20b"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3.1-pro":{"id":"moonshot/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":132000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"cohere/c4ai-aya-expanse-32b":{"id":"cohere/c4ai-aya-expanse-32b","name":"Aya Expanse 32B","description":"Open multilingual model optimized for generation across 23 languages","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-10-24","last_updated":"2024-10-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"cohere/c4ai-aya-expanse-32b"},"cohere/command-a-03-2025":{"id":"cohere/command-a-03-2025","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":288000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-03-2025"},"flexai/DeepSeek-V4-Flash-0731@eu":{"id":"flexai/DeepSeek-V4-Flash-0731@eu","name":"DeepSeek V4 Flash 0731 (FlexAI, EU)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.119,"output":0.238,"cache_read":0.0178},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"flexai/gpt-oss-120b":{"id":"flexai/gpt-oss-120b","name":"GPT OSS 120B (FlexAI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.0075},"canonical_model_id":"openai/gpt-oss-120b"},"flexai/gpt-oss-20b":{"id":"flexai/gpt-oss-20b","name":"GPT OSS 20B (FlexAI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.02,"output":0.1,"cache_read":0.003},"canonical_model_id":"openai/gpt-oss-20b"},"flexai/DeepSeek-V4.1-Flash":{"id":"flexai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash (FlexAI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.103,"output":0.715,"cache_read":0.0154},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"flexai/Step-3.7-Flash":{"id":"flexai/Step-3.7-Flash","name":"Step 3.7 Flash (FlexAI)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.03},"canonical_model_id":"stepfun/step-3.7-flash"},"flexai/DeepSeek-V4-Flash-0731":{"id":"flexai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731 (FlexAI)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.119,"output":0.238,"cache_read":0.0178},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"flexai/DeepSeek-V4.1-Flash-1017":{"id":"flexai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017 (FlexAI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.0225},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"flexai/Muse-Glimmer-30B":{"id":"flexai/Muse-Glimmer-30B","name":"Muse Glimmer 30B (FlexAI)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.045},"canonical_model_id":"meta/muse-glimmer-30b"}}},"neon":{"id":"neon","env":["NEON_AI_GATEWAY_BASE_URL","NEON_AI_GATEWAY_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"${NEON_AI_GATEWAY_BASE_URL}/v1","name":"Neon","doc":"https://neon.com/docs","models":{"gpt-5-2":{"id":"gpt-5-2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"gemini-3-6-flash":{"id":"gemini-3-6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-5-5":{"id":"gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"gpt-5-5-pro":{"id":"gpt-5-5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"qwen35-122b-a10b":{"id":"qwen35-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":25000},"cost":{"input":0.22,"output":2.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"gpt-5-4-mini":{"id":"gpt-5-4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"llama-4-maverick":{"id":"llama-4-maverick","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"grok-4-7":{"id":"grok-4-7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"glm-5-2":{"id":"glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5-6-luna":{"id":"gpt-5-6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-5-4":{"id":"gpt-5-4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"gemini-3-7-flash":{"id":"gemini-3-7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.7-flash"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"gemma-3-12b":{"id":"gemma-3-12b","name":"Gemma 3 12B","description":"Google's open-weight Gemma 3 vision-language model for text and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.5},"canonical_model_id":"google/gemma-3-12b-it"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"gemini-3-5-flash":{"id":"gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.05,"output":0.4,"cache_read":0.01},"canonical_model_id":"openai/gpt-5-nano"},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5-3-codex":{"id":"gpt-5-3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":10000},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"meta-llama-3-3-70b-instruct":{"id":"meta-llama-3-3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gemini-3-5-flash-lite":{"id":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"meta-llama-3-1-8b-instruct":{"id":"meta-llama-3-1-8b-instruct","name":"Llama 3.1 8B Instruct","description":"Meta's compact open-weight Llama 3.1 model for fast, low-cost text generation","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.45},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-5-3-flash":{"id":"glm-5-3-flash","name":"GLM-5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":524288},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"gpt-5-4-nano":{"id":"gpt-5-4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"glm-5-3":{"id":"glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"gemini-3-1-pro":{"id":"gemini-3-1-pro","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"gemini-3-8-flash":{"id":"gemini-3-8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.8-flash"},"gemini-3-1-flash-lite":{"id":"gemini-3-1-flash-lite","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"gemini-3-flash":{"id":"gemini-3-flash","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-5-6-terra":{"id":"gpt-5-6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"gpt-6-1-sol":{"id":"gpt-6-1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"gpt-5-1":{"id":"gpt-5-1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"gpt-5-6-sol":{"id":"gpt-5-6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"${NEON_AI_GATEWAY_BASE_URL}/openai/v1","shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"openai/gpt-5.6-sol"}}},"reflection":{"id":"reflection","env":["REFLECTION_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.reflection.ai/openai/v1","name":"Reflection","doc":"https://developers.reflection.ai/models","models":{"Beam-1.0-501B-A23B":{"id":"Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"}}},"above":{"id":"above","env":["ABOVE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.above.dev/v1","name":"above.dev","doc":"https://above.dev/docs","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.5077,"output":1.0154,"cache_read":0.0042},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"glm-5.2":{"id":"glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.54,"output":4.84,"cache_read":0.154},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.165,"output":0.66,"reasoning":0.66,"cache_read":0.0033},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.726,"output":2.178,"reasoning":2.178,"cache_read":0.0242},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1692,"output":0.3385,"cache_read":0.0034},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"glm-5.2-fast":{"id":"glm-5.2-fast","name":"GLM 5.2 Fast","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.31,"output":7.26,"cache_read":0.231},"canonical_model_id":"zhipuai/glm-5.2"},"mimo-v2.6-pro-ultraspeed":{"id":"mimo-v2.6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":5.0769,"output":10.1538,"cache_read":0.0423},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen 3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2.2,"output":6.6,"cache_read":0.275},"canonical_model_id":"alibaba/qwen3.8-max"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.165,"output":0.66,"reasoning":0.66,"cache_read":0.0033},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.165,"output":0.55,"cache_read":0.0319},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"tempr":{"id":"tempr","env":["TEMPR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.temprhq.io/v1","name":"Tempr Gateway","doc":"https://temprhq.io/docs/gateway-reference.html","models":{"google/gemini-embedding-001":{"id":"google/gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"cost":{"input":0.15,"output":0},"canonical_model_id":"google/gemini-embedding-001"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-flash-lite-latest"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-flash-latest"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-embedding-2":{"id":"google/gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1},"cost":{"input":0.2,"output":0},"canonical_model_id":"google/gemini-embedding-2"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"reasoning":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-flash":{"id":"deepseek/deepseek-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"xiaomi-mimo/mimo-v2.6-pro":{"id":"xiaomi-mimo/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi-mimo/mimo-v2.6-flash":{"id":"xiaomi-mimo/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi-mimo/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi-mimo/mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"cerebras/qwen-3.8-27b":{"id":"cerebras/qwen-3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.99,"output":1.49},"canonical_model_id":"alibaba/qwen3.8-27b"},"groq/allam-2-7b":{"id":"groq/allam-2-7b","name":"ALLaM-2-7b","description":"ALLaM-2-7b instruction tuned model by SDAIA","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-01-23","last_updated":"2025-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"sdaia/allam-2-7b"},"groq/openai/gpt-oss-120b":{"id":"groq/openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"groq/openai/gpt-oss-20b":{"id":"groq/openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"groq/qwen/qwen3.8-27b":{"id":"groq/qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131042,"output":16384},"cost":{"input":0.8,"output":4,"cache_read":0.4},"canonical_model_id":"alibaba/qwen3.8-27b"},"zai/glm-5v-turbo":{"id":"zai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5v-turbo"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.7-flash":{"id":"zai/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.5-flash":{"id":"zai/glm-4.5-flash","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-5.3-flashx":{"id":"zai/glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-4.6v-flash":{"id":"zai/glm-4.6v-flash","name":"GLM-4.6V-Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-4.6v":{"id":"zai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-4-5-20250929":{"id":"anthropic/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-haiku-4-5-20251001":{"id":"anthropic/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"minimax/MiniMax-M2.7-highspeed":{"id":"minimax/MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/MiniMax-M3":{"id":"minimax/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/MiniMax-M2.7":{"id":"minimax/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"mistral/zai-glm-5-2":{"id":"mistral/zai-glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"mistral/zai-glm-5-3":{"id":"mistral/zai-glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"mistral/mistral-embed":{"id":"mistral/mistral-embed","type":"embedding","name":"Mistral Embed","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":3072},"cost":{"input":0.1,"output":0},"canonical_model_id":"mistral/mistral-embed"},"mistral/mistral-small-latest":{"id":"mistral/mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-latest"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/voxtral-small-latest":{"id":"mistral/voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"mistral/voxtral-small-latest"},"mistral/mistral-medium-latest":{"id":"mistral/mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-latest"},"mistral/mistral-small-2603":{"id":"mistral/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/mistral-medium-2604":{"id":"mistral/mistral-medium-2604","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-2604"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"}}},"lmstudio":{"id":"lmstudio","env":["LMSTUDIO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://127.0.0.1:1234/v1","name":"LMStudio","doc":"https://lmstudio.ai/models","models":{"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-20b"},"qwen/qwen3-30b-a3b-2507":{"id":"qwen/qwen3-30b-a3b-2507","name":"Qwen3 30B A3B 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3-coder-30b":{"id":"qwen/qwen3-coder-30b","name":"Qwen3 Coder 30B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"}}},"auriko":{"id":"auriko","env":["AURIKO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.auriko.ai/v1","name":"Auriko","doc":"https://docs.auriko.ai","models":{"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"minimax-m2-7-highspeed":{"id":"minimax-m2-7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"qwen-3.6-plus":{"id":"qwen-3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.1,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"}}},"jiekou":{"id":"jiekou","env":["JIEKOU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.jiekou.ai/openai","name":"Jiekou.AI","doc":"https://docs.jiekou.ai/docs/support/quickstart?utm_source=github_models.dev","models":{"gpt-5.1":{"id":"gpt-5.1","name":"gpt-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5.1"},"grok-4-0709":{"id":"grok-4-0709","name":"grok-4-0709","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-01","last_updated":"2026-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":2.85,"output":14.25}},"gemini-2.5-flash-lite-preview-06-17":{"id":"gemini-2.5-flash-lite-preview-06-17","name":"gemini-2.5-flash-lite-preview-06-17","description":"First preview of Gemini 2.5 Flash-Lite, the low-cost, low-latency 2.5 model","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.09,"output":0.36},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-06-17"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"gemini-3-flash-preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.475,"output":2.85,"cache_read":0.0475},"canonical_model_id":"google/gemini-3-flash-preview"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"gpt-5.1-codex-max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5.1-codex-max"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"claude-opus-4-5-20251101","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":65536},"cost":{"input":4.75,"output":23.75,"cache_read":0.475,"cache_write":5.9375},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"gpt-5.1-codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5.1-codex"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"grok-4-1-fast-reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.19,"output":0.475},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"claude-opus-4-1-20250805","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":13.5,"output":67.5},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gpt-5-mini":{"id":"gpt-5-mini","name":"gpt-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2375,"output":1.9,"cache_read":0.0237},"canonical_model_id":"openai/gpt-5-mini"},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"grok-4-fast-non-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","cost":{"input":0.18,"output":0.45},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"gpt-5.2-pro":{"id":"gpt-5.2-pro","name":"gpt-5.2-pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":19.95,"output":159.6},"canonical_model_id":"openai/gpt-5.2-pro"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"gemini-2.5-flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.285,"output":2.375,"cache_read":0.0285},"canonical_model_id":"google/gemini-2.5-flash"},"gpt-5-nano":{"id":"gpt-5-nano","name":"gpt-5-nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.0475,"output":0.38,"cache_read":0.0047},"canonical_model_id":"openai/gpt-5-nano"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"claude-opus-4-6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"grok-code-fast-1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":10000},"cost":{"input":0.19,"output":1.425,"cache_read":0.019},"canonical_model_id":"xai/grok-code-fast-1"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"gemini-2.5-pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"google/gemini-2.5-pro"},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.045,"output":4.18,"cache_read":0.5225},"canonical_model_id":"openai/o3-mini"},"gpt-5.2":{"id":"gpt-5.2","name":"gpt-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.6625,"output":13.3,"cache_read":0.1662},"canonical_model_id":"openai/gpt-5.2"},"claude-sonnet-4-20250514":{"id":"claude-sonnet-4-20250514","name":"claude-sonnet-4-20250514","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":2.7,"output":13.5},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"claude-sonnet-4-5-20250929","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gpt-5-pro":{"id":"gpt-5-pro","name":"gpt-5-pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":14.25,"output":114},"canonical_model_id":"openai/gpt-5-pro"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4},"canonical_model_id":"openai/o4-mini"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"gpt-5.1-codex-mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2375,"output":1.9,"cache_read":0.0237},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"grok-4-1-fast-non-reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.19,"output":0.475},"canonical_model_id":"xai/grok-4.1-fast"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"claude-haiku-4-5-20251001","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gemini-2.5-flash-lite-preview-09-2025":{"id":"gemini-2.5-flash-lite-preview-09-2025","name":"gemini-2.5-flash-lite-preview-09-2025","description":"September 2025 preview of Gemini 2.5 Flash-Lite with better instruction following and multimodal quality","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.09,"output":0.36},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-09-2025"},"gemini-2.5-flash-preview-05-20":{"id":"gemini-2.5-flash-preview-05-20","name":"gemini-2.5-flash-preview-05-20","description":"Preview snapshot of Gemini 2.5 Flash with controllable thinking budgets","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1425,"output":3.325,"cache_read":0.0285},"canonical_model_id":"google/gemini-2.5-flash-preview-05-20"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":9.5,"output":38,"cache_read":0.475},"canonical_model_id":"openai/o3"},"grok-4-fast-reasoning":{"id":"grok-4-fast-reasoning","name":"grok-4-fast-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","cost":{"input":0.18,"output":0.45},"canonical_model_id":"xai/grok-4-fast-reasoning"},"claude-opus-4-20250514":{"id":"claude-opus-4-20250514","name":"claude-opus-4-20250514","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":13.5,"output":67.5},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"gpt-5.2-codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"gemini-3-pro-preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.8,"output":10.8},"canonical_model_id":"google/gemini-3-pro-preview"},"gpt-5-codex":{"id":"gpt-5-codex","name":"gpt-5-codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"openai/gpt-5-codex"},"gemini-2.5-pro-preview-06-05":{"id":"gemini-2.5-pro-preview-06-05","name":"gemini-2.5-pro-preview-06-05","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["image","text","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.1875,"output":9.5,"cache_read":0.1187},"canonical_model_id":"google/gemini-2.5-pro-preview"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"gemini-2.5-flash-lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.095,"output":0.38,"cache_read":0.0095},"canonical_model_id":"google/gemini-2.5-flash-lite"},"gpt-5-chat-latest":{"id":"gpt-5-chat-latest","name":"gpt-5-chat-latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.125,"output":9},"canonical_model_id":"openai/gpt-5-chat-latest"},"xiaomimimo/mimo-v2-flash":{"id":"xiaomimimo/mimo-v2-flash","name":"XiaomiMiMo/MiMo-V2-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"xiaomi/mimo-v2-flash"},"deepseek/deepseek-v3-0324":{"id":"deepseek/deepseek-v3-0324","name":"DeepSeek V3 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.28,"output":1.14,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.27,"output":1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.7,"output":2.5,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":100352},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2-instruct":{"id":"moonshotai/kimi-k2-instruct","name":"Kimi K2 Instruct","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":100352},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"baidu/ernie-4.5-vl-424b-a47b":{"id":"baidu/ernie-4.5-vl-424b-a47b","name":"ERNIE 4.5 VL 424B A47B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":123000,"output":16000},"cost":{"input":0.42,"output":1.25},"canonical_model_id":"baidu/ernie-4.5-vl-424b-a47b"},"baidu/ernie-4.5-300b-a47b-paddle":{"id":"baidu/ernie-4.5-300b-a47b-paddle","name":"ERNIE 4.5 300B A47B","description":"Open ERNIE text MoE with 300B total and 47B active parameters","family":"ernie","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":123000,"output":12000},"cost":{"input":0.28,"output":1.1},"canonical_model_id":"baidu/ernie-4.5-300b-a47b"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"Minimax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"qwen/qwen3-32b-fp8":{"id":"qwen/qwen3-32b-fp8","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.1,"output":0.45},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3-235b-a22b-fp8":{"id":"qwen/qwen3-235b-a22b-fp8","name":"Qwen3 235B A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"qwen/qwen3-30b-a3b-fp8":{"id":"qwen/qwen3-30b-a3b-fp8","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.09,"output":0.45},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"qwen/qwen3-coder-next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3-235b-a22b-instruct-2507":{"id":"qwen/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.15,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3-coder-480b-a35b-instruct":{"id":"qwen/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B Thinking","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22b Thinking 2507","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":3},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"minimaxai/minimax-m1-80k":{"id":"minimaxai/minimax-m1-80k","name":"MiniMax M1","description":"Open hybrid-attention MoE reasoning model with a million-token context and an 80K thinking budget","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":40000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"minimax/minimax-m1-80k"},"zai-org/glm-4.5v":{"id":"zai-org/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8},"canonical_model_id":"zhipuai/glm-4.5v"},"zai-org/glm-4.7-flash":{"id":"zai-org/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/glm-4.7":{"id":"zai-org/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/glm-4.5":{"id":"zai-org/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.5"}}},"xiaomi-token-plan-cn":{"id":"xiaomi-token-plan-cn","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan-cn.xiaomimimo.com/v1","name":"Xiaomi Token Plan (China)","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo-v2.5-tts-voicedesign":{"id":"mimo-v2.5-tts-voicedesign","type":"speech","name":"MiMo-V2.5-TTS-VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voicedesign"},"mimo-v2.5-tts-voiceclone":{"id":"mimo-v2.5-tts-voiceclone","type":"speech","name":"MiMo-V2.5-TTS-VoiceClone","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voiceclone"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5-tts":{"id":"mimo-v2.5-tts","type":"speech","name":"MiMo-V2.5-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts"}}},"lucidquery":{"id":"lucidquery","env":["LUCIDQUERY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lucidquery.com/v1","name":"LucidQuery","doc":"https://lucidquery.com/docs","models":{"lucidnova-rf1-100b":{"id":"lucidnova-rf1-100b","name":"LucidNova RF1 100B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"nova","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-09-16","release_date":"2024-12-28","last_updated":"2025-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":120000,"output":8000},"cost":{"input":2,"output":5}},"lucidquery-agi-01-frontier":{"id":"lucidquery-agi-01-frontier","name":"AGI-01 Frontier","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-06-05","release_date":"2026-06-16","last_updated":"2026-06-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":120000},"cost":{"input":5.04675,"output":24.673}},"lucidquery-nexus-coder":{"id":"lucidquery-nexus-coder","name":"LucidQuery Nexus Coder","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"lucid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-08-01","release_date":"2025-09-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":250000,"output":60000},"cost":{"input":2,"output":5}},"lucidquery-agi-01-swift":{"id":"lucidquery-agi-01-swift","name":"AGI-01 Swift","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-06-05","release_date":"2026-06-16","last_updated":"2026-06-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":120000},"cost":{"input":2.80375,"output":16.8225}}}},"vultr":{"id":"vultr","env":["VULTR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.vultrinference.com/v1","name":"Vultr","doc":"https://api.vultrinference.com/","models":{"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.75,"output":3},"canonical_model_id":"zhipuai/glm-5.2"},"mimo-v2.6-pro-rl":{"id":"mimo-v2.6-pro-rl","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":0.8},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.6-flash-rl":{"id":"mimo-v2.6-flash-rl","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.25},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.25,"output":1},"canonical_model_id":"meta/muse-glimmer-30b"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.2,"output":0.9},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-flash-next":{"id":"qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"nemotron-3.5-content-safety":{"id":"nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"deprecated","cost":{"input":0.1,"output":0.25},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.75,"output":3},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5.x-menthol":{"id":"glm-5.x-menthol","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.4,"output":1.75},"canonical_model_id":"zhipuai/glm-5"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.35},"canonical_model_id":"zhipuai/glm-5.3-flash"},"laguna-s-2.1":{"id":"laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.09,"output":0.18},"canonical_model_id":"poolside/laguna-s-2.1"},"nemotron-3-nano-omni-30b-a3b-reasoning":{"id":"nemotron-3-nano-omni-30b-a3b-reasoning","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.25},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":1},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"gmicloud":{"id":"gmicloud","env":["GMICLOUD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.gmi-serving.com/v1","name":"GMI Cloud","doc":"https://docs.gmicloud.ai/inference-engine/api-reference/llm-api-reference","models":{"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24}},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.855,"output":3.6,"cache_read":0.144},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"Qwen/Qwen3.8-Max":{"id":"Qwen/Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"Qwen/Qwen3.8-Flash":{"id":"Qwen/Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048575,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"Qwen/Qwen3.7-Max":{"id":"Qwen/Qwen3.7-Max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.392,"output":2.784,"cache_read":0.116},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":384000},"cost":{"input":0.112,"output":0.224,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-flash"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":409600,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":409600,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":409600,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"zai-org/GLM-5.2-FP8":{"id":"zai-org/GLM-5.2-FP8","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.979,"output":3.08,"cache_read":0.182},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5-FP8":{"id":"zai-org/GLM-5-FP8","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.6,"output":1.92,"cache_read":0.12},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5.1-FP8":{"id":"zai-org/GLM-5.1-FP8","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.98,"output":3.08,"cache_read":0.182},"canonical_model_id":"zhipuai/glm-5.1"}}},"qiniu-ai":{"id":"qiniu-ai","env":["QINIU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.qnaigc.com/v1","name":"Qiniu","doc":"https://developer.qiniu.com/aitokenapi","models":{"doubao-seed-2.0-lite":{"id":"doubao-seed-2.0-lite","name":"Doubao Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.08695652,"output":0.52173913,"cache_read":0.0173913,"tiers":[{"input":0.13043478,"output":0.7826087,"cache_read":0.02608696,"tier":{"type":"context","size":32000}},{"input":0.26086957,"output":1.56521739,"cache_read":0.05217391,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"doubao-seed-1.6":{"id":"doubao-seed-1.6","name":"Doubao-Seed 1.6","description":"ByteDance Seed model for long-context reasoning, instruction following, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.11594203,"output":1.15942029,"tiers":[{"input":0.17391304,"output":2.31884058,"tier":{"type":"context","size":32000}},{"input":0.34782609,"output":3.47826087,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"qwen-turbo":{"id":"qwen-turbo","name":"Qwen-Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":4096},"status":"deprecated","cost":{"input":0.04347826,"output":0.08695652},"canonical_model_id":"alibaba/qwen-turbo"},"qwen3-30b-a3b-thinking-2507":{"id":"qwen3-30b-a3b-thinking-2507","name":"Qwen3 30b A3b Thinking 2507","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":126000,"output":32000},"status":"deprecated","cost":{"input":0.10869565,"output":1.08695652}},"claude-3.5-haiku":{"id":"claude-3.5-haiku","name":"Claude 3.5 Haiku","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07-31","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"mimo-v2-flash":{"id":"mimo-v2-flash","name":"Mimo-V2-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"xiaomi/mimo-v2-flash"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":4096},"status":"deprecated","canonical_model_id":"zhipuai/glm-4.5-air"},"qwen2.5-vl-72b-instruct":{"id":"qwen2.5-vl-72b-instruct","name":"Qwen 2.5 VL 72B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":2.31884058,"output":6.95652174},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","canonical_model_id":"openai/gpt-oss-20b"},"gemini-3.0-pro-image-preview":{"id":"gemini-3.0-pro-image-preview","type":"image","name":"Gemini 3.0 Pro Image Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-3-pro-image-preview"},"gemini-3.0-pro-preview":{"id":"gemini-3.0-pro-preview","name":"Gemini 3.0 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-3-pro-preview"},"claude-3.7-sonnet":{"id":"claude-3.7-sonnet","name":"Claude 3.7 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"status":"deprecated","canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"doubao-1.5-pro-32k":{"id":"doubao-1.5-pro-32k","name":"Doubao 1.5 Pro 32k","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":12000},"status":"deprecated","cost":{"input":0.11594203,"output":0.28985507}},"claude-4.5-haiku":{"id":"claude-4.5-haiku","name":"Claude 4.5 Haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","canonical_model_id":"anthropic/claude-haiku-4-5"},"qwen3-30b-a3b-instruct-2507":{"id":"qwen3-30b-a3b-instruct-2507","name":"Qwen3 30b A3b Instruct 2507","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.10869565,"output":0.43478261},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"claude-4.0-opus":{"id":"claude-4.0-opus","name":"Claude 4.0 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","canonical_model_id":"anthropic/claude-opus-4-0"},"qwen3-max-preview":{"id":"qwen3-max-preview","name":"Qwen3 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-05","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.86956522,"output":3.47826087,"tiers":[{"input":1.44927536,"output":5.79710145,"tier":{"type":"context","size":32000}},{"input":2.17391304,"output":8.69565217,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max-preview"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Gemini 2.5 Flash Image","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash-image"},"doubao-seed-2.0-mini":{"id":"doubao-seed-2.0-mini","name":"Doubao Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.02898551,"output":0.28985507,"cache_read":0.0057971,"tiers":[{"input":0.05797101,"output":0.57971014,"cache_read":0.0115942,"tier":{"type":"context","size":32000}},{"input":0.11594203,"output":1.15942029,"cache_read":0.02318841,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash"},"claude-4.5-sonnet":{"id":"claude-4.5-sonnet","name":"Claude 4.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","canonical_model_id":"anthropic/claude-sonnet-4-5"},"glm-4.5":{"id":"glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"zhipuai/glm-4.5"},"claude-4.5-opus":{"id":"claude-4.5-opus","name":"Claude 4.5 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":200000},"status":"deprecated","canonical_model_id":"anthropic/claude-opus-4-5"},"deepseek-v3.1":{"id":"deepseek-v3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":1.73913043},"canonical_model_id":"deepseek/deepseek-v3.1"},"qwen-vl-max-2025-01-25":{"id":"qwen-vl-max-2025-01-25","name":"Qwen VL-MAX-2025-01-25","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.43478261,"output":1.30434783}},"doubao-seed-1.6-thinking":{"id":"doubao-seed-1.6-thinking","name":"Doubao-Seed 1.6 Thinking","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-15","last_updated":"2025-08-15","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.11594203,"output":1.15942029,"tiers":[{"input":0.17391304,"output":2.31884058,"tier":{"type":"context","size":32000}},{"input":0.34782609,"output":3.47826087,"tier":{"type":"context","size":128000}}]}},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.14492754,"output":0.57971014},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-2.5-pro"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235b A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"doubao-seed-2.0-code":{"id":"doubao-seed-2.0-code","name":"Doubao Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.46376812,"output":2.31884058,"cache_read":0.09275362,"tiers":[{"input":0.69565217,"output":3.47826087,"cache_read":0.13913043,"tier":{"type":"context","size":32000}},{"input":1.39130435,"output":6.95652174,"cache_read":0.27826087,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"doubao-seed-1.6-flash":{"id":"doubao-seed-1.6-flash","name":"Doubao-Seed 1.6 Flash","description":"Low-latency ByteDance Seed model for high-throughput chat, extraction, and lightweight tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.02173913,"output":0.2173913,"tiers":[{"input":0.04347826,"output":0.43478261,"tier":{"type":"context","size":32000}},{"input":0.08695652,"output":0.86956522,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.17391304,"output":1.04347826,"tiers":[{"input":0.43478261,"output":2.60869565,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":4096},"status":"deprecated","cost":{"input":0.86956522,"output":3.47826087,"tiers":[{"input":1.30434783,"output":5.2173913,"tier":{"type":"context","size":32000}},{"input":2.17391304,"output":8.69565217,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.86956522,"output":3.47826087,"tiers":[{"input":1.44927536,"output":5.79710145,"tier":{"type":"context","size":32000}},{"input":2.17391304,"output":8.69565217,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"MiniMax-M1":{"id":"MiniMax-M1","name":"MiniMax M1","description":"Open hybrid-attention MoE reasoning model with a million-token context and an 80K thinking budget","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":80000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"minimax/minimax-m1-80k"},"deepseek-r1-0528":{"id":"deepseek-r1-0528","name":"DeepSeek-R1-0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"deepseek/deepseek-r1-0528"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":4096},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"alibaba/qwen3-32b"},"qwen3-next-80b-a3b-thinking":{"id":"qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B Thinking","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.14492754,"output":1.44927536},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"claude-3.5-sonnet":{"id":"claude-3.5-sonnet","name":"Claude 3.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","canonical_model_id":"anthropic/claude-3-5-sonnet-20241022"},"gemini-2.0-flash":{"id":"gemini-2.0-flash","name":"Gemini 2.0 Flash","description":"Earlier Gemini Flash workhorse for responsive multimodal apps and tool use","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-2.0-flash"},"qwen3-235b-a22b-thinking-2507":{"id":"qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22B Thinking 2507","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"status":"deprecated","cost":{"input":0.28985507,"output":2.89855072},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen3-vl-30b-a3b-thinking":{"id":"qwen3-vl-30b-a3b-thinking","name":"Qwen3-Vl 30b A3b Thinking","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-09","last_updated":"2026-02-09","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.10869565,"output":1.08695652}},"kling-v2-6":{"id":"kling-v2-6","type":"video","name":"Kling-V2 6","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-13","last_updated":"2026-01-13","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":99999999,"output":99999999}},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen 3 235B A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"gemini-3.0-flash-preview":{"id":"gemini-3.0-flash-preview","name":"Gemini 3.0 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-3-flash-preview"},"gemini-2.0-flash-lite":{"id":"gemini-2.0-flash-lite","name":"Gemini 2.0 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":8192},"status":"deprecated","canonical_model_id":"google/gemini-2.0-flash-lite"},"deepseek-v3":{"id":"deepseek-v3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"status":"deprecated","cost":{"input":0.28985507,"output":1.15942029},"canonical_model_id":"deepseek/deepseek-v3"},"doubao-seed-2.0-pro":{"id":"doubao-seed-2.0-pro","name":"Doubao Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.46376812,"output":2.31884058,"cache_read":0.09275362,"tiers":[{"input":0.69565217,"output":3.47826087,"cache_read":0.13913043,"tier":{"type":"context","size":32000}},{"input":1.39130435,"output":6.95652174,"cache_read":0.27826087,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":4096},"status":"deprecated","cost":{"input":0.10869565,"output":0.43478261},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"claude-4.1-opus":{"id":"claude-4.1-opus","name":"Claude 4.1 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","canonical_model_id":"anthropic/claude-opus-4-1"},"kimi-k2":{"id":"kimi-k2","name":"Kimi K2","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"moonshotai/kimi-k2"},"doubao-1.5-vision-pro":{"id":"doubao-1.5-vision-pro","name":"Doubao 1.5 Vision Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"status":"deprecated","cost":{"input":0.43478261,"output":1.30434783}},"qwen-max-2025-01-25":{"id":"qwen-max-2025-01-25","name":"Qwen2.5-Max-2025-01-25","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.34782609,"output":1.39130435},"canonical_model_id":"alibaba/qwen-max"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":64000},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash-lite"},"claude-4.0-sonnet":{"id":"claude-4.0-sonnet","name":"Claude 4.0 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","canonical_model_id":"anthropic/claude-sonnet-4-0"},"deepseek-r1":{"id":"deepseek-r1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"deepseek/deepseek-r1"},"doubao-1.5-thinking-pro":{"id":"doubao-1.5-thinking-pro","name":"Doubao 1.5 Thinking Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058}},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"Z-AI/GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":200000},"status":"deprecated","cost":{"input":1.04347826,"output":1.82608696},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"Z-Ai/GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":200000},"status":"deprecated","cost":{"input":0.45913043,"output":1.81565217},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"Z-Ai/GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.57971014,"output":2.60869565,"tiers":[{"input":0.86956522,"output":3.1884058,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"z-ai/autoglm-phone-9b":{"id":"z-ai/autoglm-phone-9b","name":"Z-Ai/Autoglm Phone 9b","description":"GLM vision model for visual reasoning, documents, and multimodal agents","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":12800,"output":4096},"status":"deprecated"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Stepfun/Step-3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"input":256000,"output":4096},"status":"deprecated","canonical_model_id":"stepfun/step-3.5-flash"},"meituan/longcat-flash-lite":{"id":"meituan/longcat-flash-lite","name":"Meituan/Longcat-Flash-Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-06","last_updated":"2026-02-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":320000},"status":"deprecated","cost":{"input":0,"output":0}},"meituan/longcat-flash-chat":{"id":"meituan/longcat-flash-chat","name":"Meituan/Longcat-Flash-Chat","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-11-05","last_updated":"2025-11-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"status":"deprecated"},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek/DeepSeek-V3.2-Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":0.43478261}},"deepseek/deepseek-math-v2":{"id":"deepseek/deepseek-math-v2","name":"Deepseek/Deepseek-Math-V2","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-04","last_updated":"2025-12-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":160000,"output":160000},"status":"deprecated"},"deepseek/deepseek-v3.1-terminus-thinking":{"id":"deepseek/deepseek-v3.1-terminus-thinking","name":"DeepSeek/DeepSeek-V3.1-Terminus-Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":1.73913043}},"deepseek/deepseek-v3.1-terminus":{"id":"deepseek/deepseek-v3.1-terminus","name":"DeepSeek/DeepSeek-V3.1-Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.57971014,"output":1.73913043}},"deepseek/deepseek-v3.2-251201":{"id":"deepseek/deepseek-v3.2-251201","name":"Deepseek/DeepSeek-V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":0.43478261,"cache_read":0.02898551},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v3.2-exp-thinking":{"id":"deepseek/deepseek-v3.2-exp-thinking","name":"DeepSeek/DeepSeek-V3.2-Exp-Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"status":"deprecated","cost":{"input":0.28985507,"output":0.43478261}},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":100000},"status":"deprecated","canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Moonshotai/Kimi-K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0.57971014,"output":3.04347826},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":100000},"status":"deprecated","cost":{"input":0.57971014,"output":2.31884058},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"OpenAI/GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5":{"id":"openai/gpt-5","name":"OpenAI/GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","canonical_model_id":"openai/gpt-5"},"stepfun-ai/gelab-zero-4b-preview":{"id":"stepfun-ai/gelab-zero-4b-preview","name":"Stepfun-Ai/Gelab Zero 4b Preview","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":4096},"status":"deprecated"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"Minimax/Minimax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"status":"deprecated","cost":{"input":0.30434783,"output":1.2173913},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"Minimax/Minimax-M2.5 Highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"status":"deprecated","cost":{"input":0.60869565,"output":2.43478261},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"Minimax/Minimax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"cost":{"input":0.30434783,"output":1.2173913},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"Minimax/Minimax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":0.30434783,"output":1.2173913},"canonical_model_id":"minimax/MiniMax-M2"},"x-ai/grok-4.1-fast":{"id":"x-ai/grok-4.1-fast","name":"x-AI/Grok-4.1-Fast","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated"},"x-ai/grok-4.1-fast-reasoning":{"id":"x-ai/grok-4.1-fast-reasoning","name":"X-Ai/Grok 4.1 Fast Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":20000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast-reasoning"},"x-ai/grok-4-fast-non-reasoning":{"id":"x-ai/grok-4-fast-non-reasoning","name":"X-Ai/Grok-4-Fast-Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4-fast-non-reasoning"},"x-ai/grok-code-fast-1":{"id":"x-ai/grok-code-fast-1","name":"x-AI/Grok-Code-Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":10000},"status":"deprecated","canonical_model_id":"xai/grok-code-fast-1"},"x-ai/grok-4.1-fast-non-reasoning":{"id":"x-ai/grok-4.1-fast-non-reasoning","name":"X-Ai/Grok 4.1 Fast Non Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast"},"x-ai/grok-4-fast-reasoning":{"id":"x-ai/grok-4-fast-reasoning","name":"X-Ai/Grok-4-Fast-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated","canonical_model_id":"xai/grok-4-fast-reasoning"},"x-ai/grok-4-fast":{"id":"x-ai/grok-4-fast","name":"x-AI/Grok-4-Fast","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-20","last_updated":"2025-09-20","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"status":"deprecated"}}},"agentrouter":{"id":"agentrouter","env":["AGENTROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://agentrouter.org/v1","name":"AgentRouter","doc":"https://agentrouter.org/docs/opencode.html","models":{"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://agentrouter.org/v1"},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://agentrouter.org/v1"},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-5.6-sol"}}},"zai-coding-plan":{"id":"zai-coding-plan","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.z.ai/api/coding/paas/v4","name":"Z.AI Coding Plan","doc":"https://docs.z.ai/devpack/overview","models":{"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"glm-5.2-highspeed":{"id":"glm-5.2-highspeed","name":"GLM-5.2 Highspeed","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5.3-highspeed":{"id":"glm-5.3-highspeed","name":"GLM-5.3 Highspeed","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5-turbo":{"id":"glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"ebcloud":{"id":"ebcloud","env":["EBCLOUD_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://maas-api.ebcloud.com/v1","name":"EBCloud","doc":"https://docs.ebtech.com/ai/model-api.html","models":{"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.9286,"output":3.8571},"canonical_model_id":"moonshotai/kimi-k2.6"},"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.4286,"output":0.8571},"canonical_model_id":"deepseek/deepseek-v4-pro"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.143,"output":0.2857},"canonical_model_id":"deepseek/deepseek-v4-flash"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.8571,"output":3.4286},"canonical_model_id":"zhipuai/glm-5.1"}}},"deepseek":{"id":"deepseek","env":["DEEPSEEK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.deepseek.com","name":"DeepSeek","doc":"https://api-docs.deepseek.com/quick_start/pricing","models":{"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.66,"output":1.98,"reasoning":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-flash":{"id":"deepseek-flash","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"status":"deprecated","cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"status":"deprecated","cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"}}},"modal":{"id":"modal","env":["MODAL_PROXY_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.us-west.modal.direct/v1","name":"Modal","doc":"https://modal.com/docs/guide/endpoints","models":{"nvidia/GLM-5.3-Flash-NVFP4":{"id":"nvidia/GLM-5.3-Flash-NVFP4","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"thinkingmachines/Inkling-NVFP4":{"id":"thinkingmachines/Inkling-NVFP4","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.2,"output":5,"cache_read":0.27},"canonical_model_id":"thinkingmachines/inkling"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"reasoning":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8-Max","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.45,"output":1.5,"cache_read":0.09},"canonical_model_id":"zhipuai/glm-5.3-flash"},"meta-models/Muse-Spark-Open-816B-A42B":{"id":"meta-models/Muse-Spark-Open-816B-A42B","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"}}},"oci":{"id":"oci","env":["OCI_GENAI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.generativeai.us-chicago-1.oci.oraclecloud.com/openai/v1","name":"OCI Generative AI","doc":"https://docs.oracle.com/en-us/iaas/Content/generative-ai/pretrained-models.htm","models":{"openai.gpt-oss-120b":{"id":"openai.gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"meta.llama-4-maverick-17b-128e-instruct-fp8":{"id":"meta.llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":16384},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"xai.grok-4.6":{"id":"xai.grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai.grok-4.3":{"id":"xai.grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai.grok-4.20-non-reasoning":{"id":"xai.grok-4.20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai.grok-4.20-reasoning":{"id":"xai.grok-4.20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"meta.llama-3.3-70b-instruct":{"id":"meta.llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta.llama-4-scout-17b-16e-instruct":{"id":"meta.llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":16384},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"openai.gpt-oss-20b":{"id":"openai.gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"}}},"moonshotai":{"id":"moonshotai","env":["MOONSHOT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.moonshot.ai/v1","name":"Moonshot AI","doc":"https://platform.moonshot.ai/docs/api/chat","models":{"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"kimi-k3.1-pro-highspeed":{"id":"kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code HighSpeed","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code"}}},"volcengine-coding-plan":{"id":"volcengine-coding-plan","env":["ARK_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ark.cn-beijing.volces.com/api/coding/v3","name":"Volcengine Ark Coding Plan","doc":"https://www.volcengine.com/docs/82379/1928261","models":{"doubao-seed-2.0-lite":{"id":"doubao-seed-2.0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"doubao-seed-2.1-turbo":{"id":"doubao-seed-2.1-turbo","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k3"},"doubao-seed-evolving":{"id":"doubao-seed-evolving","name":"Seed Evolving","description":"Rolling ByteDance Seed model for rapidly updated reasoning, coding, and agent capabilities","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"bytedance-seed/seed-evolving"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"engy":{"id":"engy","env":["ENGY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.engy.ai/v1","name":"engy","doc":"https://engy.ai/pricing","models":{"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":229376,"output":32768},"cost":{"input":0.68,"output":1.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":983040,"output":65536},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":983040,"output":65536},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1113088,"input":1047552,"output":65536},"cost":{"input":1.95,"output":9.75,"cache_read":0.195},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":920576,"output":128000},"cost":{"input":0.045,"output":0.09,"cache_read":0.009},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"input":294912,"output":32768},"cost":{"input":0.98,"output":3.08,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":208192,"input":200000,"output":8192},"cost":{"input":0.045,"output":0.3,"cache_read":0.015},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"input":294912,"output":32768},"cost":{"input":0.135,"output":0.45,"cache_read":0.027},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1001536,"input":936000,"output":65536},"cost":{"input":0.045,"output":0.32,"cache_read":0.015},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"greenpt":{"id":"greenpt","env":["GREENPT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.greenpt.ai/v1","name":"GreenPT","doc":"https://docs.greenpt.ai","models":{"gemma-3-27b-it":{"id":"gemma-3-27b-it","name":"Gemma 3 27B","description":"Google Gemma 3 multimodal model for chat, reasoning, and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":40000,"output":8192},"status":"deprecated","cost":{"input":0.342,"output":0.684},"canonical_model_id":"google/gemma-3-27b-it"},"glm-5.2-caveman-ultra":{"id":"glm-5.2-caveman-ultra","name":"GLM-5.2 Caveman Ultra","description":"glm-5.2 carrying a built-in ruleset that compresses prose, keeping code and technical detail verbatim. The most aggressive tier, close to answer-only. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.863555,"output":4.317775,"cache_read":0.222057},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.228,"output":0.798},"canonical_model_id":"openai/gpt-oss-120b"},"mistral-medium-3.5-128b":{"id":"mistral-medium-3.5-128b","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":2.052,"output":10.26},"canonical_model_id":"mistral/mistral-medium-2604"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":16384},"cost":{"input":1.254,"output":1.254},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2-ponytail":{"id":"glm-5.2-ponytail","name":"GLM-5.2 Ponytail","description":"glm-5.2 carrying a built-in ruleset that compresses generated code, preferring platform features over custom code. The middle tier, and the ruleset as its authors wrote it. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":0.285,"output":1.083},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.255552,"output":1.27776,"cache_read":0.0127776},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"glm-5.2-caveman":{"id":"glm-5.2-caveman","name":"GLM-5.2 Caveman","description":"glm-5.2 carrying a built-in ruleset that compresses prose, keeping code and technical detail verbatim. The middle tier, and the ruleset as its authors wrote it. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"status":"deprecated","cost":{"input":1.756,"output":5.518},"canonical_model_id":"zhipuai/glm-5.1"},"green-r":{"id":"green-r","name":"Green R","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.399,"output":1.083},"canonical_model_id":"openai/gpt-oss-120b"},"green-l-raw":{"id":"green-l-raw","name":"Green L Raw","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.285,"output":0.912},"canonical_model_id":"mistral/mistral-small-2506"},"voxtral-small-24b-2507":{"id":"voxtral-small-24b-2507","name":"Voxtral Small 24B","description":"Mistral Voxtral audio-understanding model for speech and transcription tasks","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"status":"deprecated","cost":{"input":0.228,"output":0.513},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"glm-5.2-caveman-lite":{"id":"glm-5.2-caveman-lite","name":"GLM-5.2 Caveman Lite","description":"glm-5.2 carrying a built-in ruleset that compresses prose, keeping code and technical detail verbatim. The gentlest tier: it cuts filler only and keeps the explanation intact. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"green-s-pro":{"id":"green-s-pro","type":"transcription","name":"Green S Pro","description":"GreenPT advanced speech-to-text model with multilingual transcription support","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-02","last_updated":"2025-02","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":8192},"provider":{"api":"https://api.greenpt.ai"}},"glm-5.2-ponytail-lite":{"id":"glm-5.2-ponytail-lite","name":"GLM-5.2 Ponytail Lite","description":"glm-5.2 carrying a built-in ruleset that compresses generated code, preferring platform features over custom code. The gentlest tier: it cuts filler only and keeps the explanation intact. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2-ponytail-ultra":{"id":"glm-5.2-ponytail-ultra","name":"GLM-5.2 Ponytail Ultra","description":"glm-5.2 carrying a built-in ruleset that compresses generated code, preferring platform features over custom code. The most aggressive tier, close to answer-only. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"pixtral-12b-2409":{"id":"pixtral-12b-2409","name":"Pixtral 12B","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-09-11","last_updated":"2024-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.285,"output":0.285},"canonical_model_id":"mistral/pixtral-12b"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3.762,"output":18.81,"cache_read":0.9405},"canonical_model_id":"moonshotai/kimi-k3"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen3 235B MoE instruct model for long-context multilingual chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":1.026,"output":3.078},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"kimi-k2.6-fast":{"id":"kimi-k2.6-fast","name":"Kimi K2.6 Fast","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":1.655,"output":8.778},"canonical_model_id":"moonshotai/kimi-k2.6"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.7524,"output":4.275,"cache_read":0.2508},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.1596,"output":0.399,"cache_read":0.0456},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.798,"output":4.959},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"glm-5.2-honey":{"id":"glm-5.2-honey","name":"GLM-5.2 Honey","description":"glm-5.2 carrying a built-in ruleset that compresses both generated code and prose. The middle tier, and the ruleset as its authors wrote it. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2-honey-lite":{"id":"glm-5.2-honey-lite","name":"GLM-5.2 Honey Lite","description":"glm-5.2 carrying a built-in ruleset that compresses both generated code and prose. The gentlest tier: it cuts filler only and keeps the explanation intact. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.27754,"output":5.11016,"cache_read":0.319385},"canonical_model_id":"zhipuai/glm-5.3"},"holo2-30b-a3b":{"id":"holo2-30b-a3b","name":"Holo2 30B A3B","description":"H Company Holo2 vision model for GUI navigation and computer-use agents","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11","last_updated":"2025-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":22016,"output":16384},"status":"deprecated","cost":{"input":0.399,"output":0.969}},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.49346,"output":2.4673,"cache_read":0.123365},"canonical_model_id":"minimax/MiniMax-M2.5"},"devstral-2-123b-instruct-2512":{"id":"devstral-2-123b-instruct-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":16384},"status":"deprecated","cost":{"input":0.57,"output":2.736},"canonical_model_id":"mistral/devstral-2512"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.342,"output":2.052},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.127754,"output":0.511016,"cache_read":0.0255508},"canonical_model_id":"zhipuai/glm-5.3-flash"},"green-r-raw":{"id":"green-r-raw","name":"Green R Raw","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.399,"output":1.083},"canonical_model_id":"openai/gpt-oss-120b"},"green-s":{"id":"green-s","type":"transcription","name":"Green S","description":"GreenPT speech-to-text model for pre-recorded and live transcription","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":8192},"provider":{"api":"https://api.greenpt.ai"}},"green-l":{"id":"green-l","name":"Green L","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.285,"output":0.912},"canonical_model_id":"mistral/mistral-small-2506"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.228,"output":0.456},"canonical_model_id":"mistral/mistral-small-2506"},"gemma4":{"id":"gemma4","name":"gemma4","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.57,"output":1.71},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"glm-5.2-honey-ultra":{"id":"glm-5.2-honey-ultra","name":"GLM-5.2 Honey Ultra","description":"glm-5.2 carrying a built-in ruleset that compresses both generated code and prose. The most aggressive tier, close to answer-only. Same upstream model and price per token as glm-5.2, with fewer output tokens.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.738325,"output":5.1589,"cache_read":0.4486},"canonical_model_id":"zhipuai/glm-5.2"}}},"opencode-go":{"id":"opencode-go","env":["OPENCODE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://opencode.ai/inference/go/openai/v1","name":"OpenCode Go","doc":"https://opencode.ai/v2/docs/console/go/","models":{"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for coding and agentic workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen3.7-plus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.4,"output":1.6,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro (New)","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"step-5-preview-free":{"id":"step-5-preview-free","name":"Step 5 Preview Free","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"stepfun/step-5-preview"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax-m3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax-m3.1-flash-free":{"id":"minimax-m3.1-flash-free","name":"MiniMax M3.1 Flash Free","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"beam-free":{"id":"beam-free","name":"Beam Free","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo V2.5","description":"MiMo omni model for text, image, video, audio, and agents","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"longcat-2.0":{"id":"longcat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.0"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"cat-jam-free":{"id":"cat-jam-free","name":"Cat Jam Free","description":"A text-only reasoning model for coding and agentic workflows, with tool calling and structured JSON outputs.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"space-bunny":{"id":"space-bunny","name":"Space Bunny","description":"Anonymous preview reasoning model for coding, agentic tasks, tool use, and multimodal input","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":524288,"output":524288},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.03,"cache_write":0}},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"muse-spark-1.2-contributor":{"id":"muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"longcat-2.5-preview-free":{"id":"longcat-2.5-preview-free","name":"LongCat 2.5 Preview Free","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meituan/longcat-2.5-preview"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter multimodal flagship for coding, professional work, and long-horizon agentic workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/go/anthropic/v1"},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/go/openai/v1"},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-4-argon-contributor":{"id":"gemini-4-argon-contributor","name":"Gemini 4 Argon Contributor","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/go/google/v1beta"},"cost":{"input":4,"output":20,"cache_read":0.2,"input_audio":4},"canonical_model_id":"google/gemini-4-argon"}}},"snowflake-cortex":{"id":"snowflake-cortex","env":["SNOWFLAKE_ACCOUNT","SNOWFLAKE_CORTEX_PAT"],"npm":"@ai-sdk/openai-compatible","api":"https://${SNOWFLAKE_ACCOUNT}.snowflakecomputing.com/api/v2/cortex/v1","name":"Snowflake Cortex","doc":"https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-rest-api","models":{"openai-gpt-5.6-terra":{"id":"openai-gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai-gpt-5-nano":{"id":"openai-gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":0.06,"output":0.44,"cache_read":0.01},"canonical_model_id":"openai/gpt-5-nano"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"openai-gpt-5-mini":{"id":"openai-gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5-mini"},"openai-gpt-5":{"id":"openai-gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"openai-gpt-5.5":{"id":"openai-gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"beta","cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"openai-gpt-5.6-sol":{"id":"openai-gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"snowflake-llama3.3-70b":{"id":"snowflake-llama3.3-70b","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"openai-gpt-5.6-luna":{"id":"openai-gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gemini-3.1-pro":{"id":"gemini-3.1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"beta","canonical_model_id":"google/gemini-3.1-pro-preview"},"openai-gpt-4.1":{"id":"openai-gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai-gpt-5.2":{"id":"openai-gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.2"},"openai-gpt-5.1":{"id":"openai-gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"input":272000,"output":8192},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"mistral-large2":{"id":"mistral-large2","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/mistral-large-2407"},"openai-gpt-5.4":{"id":"openai-gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"deepseek-r1":{"id":"deepseek-r1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"canonical_model_id":"deepseek/deepseek-r1"}}},"meta":{"id":"meta","env":["META_MODEL_API_KEY"],"npm":"@ai-sdk/openai","api":"https://api.meta.ai/v1","name":"Meta","doc":"https://dev.meta.ai/docs","models":{"muse-spark-1.3-contributor":{"id":"muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"muse-spark-1.2-contributor":{"id":"muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"}}},"kenari":{"id":"kenari","env":["KENARI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://kenari.id/v1","name":"Kenari","doc":"https://kenari.id/docs","models":{"gemini-3-6-flash":{"id":"gemini-3-6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.34285714,"output":1.71428571,"cache_read":0.03428571,"cache_write":0.01714286},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-5-5":{"id":"gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5.71428571,"output":36,"cache_read":0.57142857},"canonical_model_id":"openai/gpt-5.5"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.10285714,"output":0.40571429,"cache_read":0.04114286},"canonical_model_id":"google/gemma-4-31b-it"},"mimo-v2-5-pro":{"id":"mimo-v2-5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.52,"output":1.02857143},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"kimi-k2-6:free":{"id":"kimi-k2-6:free","name":"Kimi K2.6 (Free)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-flash:free":{"id":"deepseek-v4-flash:free","name":"DeepSeek V4 Flash (Free)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.036,"output":0.2,"cache_read":0.0036},"canonical_model_id":"openai/gpt-oss-120b"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57142857,"output":2.85714286,"cache_read":0.28571429},"canonical_model_id":"moonshotai/kimi-k2.6"},"step-3-7-flash:free":{"id":"step-3-7-flash:free","name":"Step 3.7 Flash (Free)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"stepfun/step-3.7-flash"},"gpt-5-4-mini":{"id":"gpt-5-4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.85714286,"output":5.37142857},"canonical_model_id":"openai/gpt-5.4-mini"},"nano-banana-2":{"id":"nano-banana-2","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"canonical_model_id":"google/gemini-3.1-flash-image"},"grok-imagine-image-2-0":{"id":"grok-imagine-image-2-0","type":"image","name":"Grok Imagine Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":64000,"output":0},"canonical_model_id":"xai/grok-imagine-image-2.0"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.57142857,"output":1.14285714,"cache_read":0.00571429},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.036,"output":0.15428571,"cache_read":0.00371429},"canonical_model_id":"openai/gpt-oss-20b"},"glm-5-2":{"id":"glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.28571429,"output":3.42857143,"cache_read":0.05314286},"canonical_model_id":"zhipuai/glm-5.2"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.36,"output":1.42857143,"cache_read":0.06857143},"canonical_model_id":"minimax/MiniMax-M2.7"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.42857143,"output":7.14285714,"cache_read":0.57142857,"cache_write":1.71428571},"canonical_model_id":"anthropic/claude-sonnet-5"},"kimi-k2-7-code:free":{"id":"kimi-k2-7-code:free","name":"Kimi K2.7 Code (Free)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-5-6-luna":{"id":"gpt-5-6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.2,"output":1.14285714,"cache_read":0.02,"cache_write":0.25714286},"canonical_model_id":"openai/gpt-5.6-luna"},"grok-imagine-image":{"id":"grok-imagine-image","type":"image","name":"Grok Imagine Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"canonical_model_id":"xai/grok-imagine-image"},"gemini-3-7-flash":{"id":"gemini-3-7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.34285714,"output":1.71428571,"cache_read":0.03428571,"cache_write":0.01714286},"canonical_model_id":"google/gemini-3.7-flash"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.36,"output":1.42857143,"cache_read":0.06857143},"canonical_model_id":"minimax/MiniMax-M3"},"step-3-7-flash":{"id":"step-3-7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.24,"output":1.37142857,"cache_read":0.048},"canonical_model_id":"stepfun/step-3.7-flash"},"nemotron-3-nano-30b-a3b":{"id":"nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.71428571,"output":28.57142857,"cache_read":2.28571429,"cache_write":5.71428571},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.6,"output":17.71428571,"cache_read":0.36,"cache_write":4.45714286},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"kimi-k2-7-code":{"id":"kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.57142857,"output":2.85714286,"cache_read":0.11428571},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15714286,"output":0.31428571,"cache_read":0.00371429},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"minimax-m2-7-highspeed":{"id":"minimax-m2-7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.59428571,"output":2.4,"cache_read":0.11428571},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"grok-4-5":{"id":"grok-4-5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.28571429,"output":6.85714286,"cache_read":0.34285714},"canonical_model_id":"xai/grok-4.5"},"mimo-v2-5":{"id":"mimo-v2-5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.16571429,"output":0.33142857},"canonical_model_id":"xiaomi/mimo-v2.5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":12,"output":57.14285714,"cache_read":1.2,"cache_write":14.85714286},"canonical_model_id":"anthropic/claude-fable-5"},"gemini-3-5-flash":{"id":"gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-3.5-flash"},"glm-5-1":{"id":"glm-5-1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"status":"deprecated","canonical_model_id":"zhipuai/glm-5.1"},"grok-build-0-1":{"id":"grok-build-0-1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"status":"deprecated","canonical_model_id":"xai/grok-build-0.1"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.85714286,"output":8.57142857,"cache_read":0.11428571},"canonical_model_id":"moonshotai/kimi-k3"},"glm-4-7-flash:free":{"id":"glm-4-7-flash:free","name":"GLM-4.7-Flash (Free)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"mimo-v2-5:free":{"id":"mimo-v2-5:free","name":"MiMo-V2.5 (Free)","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.71428571,"output":29.71428571,"cache_read":0.57142857,"cache_write":7.42857143},"canonical_model_id":"anthropic/claude-opus-4-8"},"qwen3-7-plus":{"id":"qwen3-7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.38285714,"output":1.48571429,"cache_read":0.07428571,"cache_write":0.48},"canonical_model_id":"alibaba/qwen3.7-plus"},"glm-5-3-flash":{"id":"glm-5-3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11428571,"output":0.34285714,"cache_read":0.02285714},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-3-1-flash-tts":{"id":"gemini-3-1-flash-tts","type":"speech","name":"Gemini 3.1 Flash TTS Preview","description":"Low-latency speech generation with steerable prompts and expressive audio tags","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-15","last_updated":"2026-04-15","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"provider":{"npm":"@ai-sdk/openai"},"canonical_model_id":"google/gemini-3.1-flash-tts-preview"},"grok-4-6":{"id":"grok-4-6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.28571429,"output":6.85714286,"cache_read":0.34285714},"canonical_model_id":"xai/grok-4.6"},"nano-banana-pro":{"id":"nano-banana-pro","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"canonical_model_id":"google/gemini-3-pro-image"},"glm-5-3":{"id":"glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.28571429,"output":3.42857143,"cache_read":0.05314286},"canonical_model_id":"zhipuai/glm-5.3"},"gemini-3-1-pro":{"id":"gemini-3-1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.2,"output":7.14285714,"cache_read":0.12,"cache_write":0.22285714},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini-2-5-flash-lite":{"id":"gemini-2-5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash-lite"},"whisper-large-v3-turbo":{"id":"whisper-large-v3-turbo","type":"transcription","name":"Whisper Large v3 Turbo","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"provider":{"npm":"@ai-sdk/openai"},"canonical_model_id":"openai/whisper-large-v3-turbo"},"gemini-2-5-flash":{"id":"gemini-2-5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","canonical_model_id":"google/gemini-2.5-flash"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.15428571,"output":0.62857143,"cache_read":0.01542857},"canonical_model_id":"tencent/hy3"},"qwen3-8-max":{"id":"qwen3-8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.4,"output":6.85714286,"cache_read":0.29714286,"cache_write":2.97142857},"canonical_model_id":"alibaba/qwen3.8-max"},"mistral-medium-3-5:free":{"id":"mistral-medium-3-5:free","name":"Mistral Medium 3.5 (Free)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-medium-2604"},"nemotron-3-super-120b-a12b":{"id":"nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.09714286,"output":0.48,"cache_read":0.00971429},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"gemini-3-1-flash-lite":{"id":"gemini-3-1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.28571429,"output":1.71428571,"cache_read":0.02857143,"cache_write":0.08571429},"canonical_model_id":"google/gemini-3.1-flash-lite"},"hy3:free":{"id":"hy3:free","name":"Hy3 (Free)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"tencent/hy3"},"nemotron-3-ultra-550b-a55b":{"id":"nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.68571429,"output":4.28571429,"cache_read":0.06857143},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nano-banana-2-lite":{"id":"nano-banana-2-lite","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.71428571,"output":29.71428571,"cache_read":0.57142857,"cache_write":7.42857143},"canonical_model_id":"anthropic/claude-opus-4-7"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15714286,"output":0.31428571,"cache_read":0.00371429},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":272000,"output":16384},"canonical_model_id":"openai/gpt-image-2"},"gpt-5-6-terra":{"id":"gpt-5-6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"status":"deprecated","cost":{"input":2.4,"output":14.28571429},"canonical_model_id":"openai/gpt-5.6-terra"},"nemotron-3-super-120b-a12b:free":{"id":"nemotron-3-super-120b-a12b:free","name":"Nemotron 3 Super 120B A12B (Free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"mistral-large:free":{"id":"mistral-large:free","name":"Mistral Large (Free)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-large-latest"},"gpt-5-6-sol":{"id":"gpt-5-6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.97142857,"output":17.71428571,"cache_read":1.18857143,"cache_write":3.71428571},"canonical_model_id":"openai/gpt-5.6-sol"}}},"tencent-coding-plan":{"id":"tencent-coding-plan","env":["TENCENT_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lkeap.cloud.tencent.com/coding/v3","name":"Tencent Coding Plan (China)","doc":"https://cloud.tencent.com/document/product/1772/128947","models":{"hunyuan-turbos":{"id":"hunyuan-turbos","name":"Hunyuan-TurboS","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"hunyuan-2.0-thinking":{"id":"hunyuan-2.0-thinking","name":"Tencent HY 2.0 Think","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"hunyuan-2.0-instruct":{"id":"hunyuan-2.0-instruct","name":"Tencent HY 2.0 Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"tc-code-latest":{"id":"tc-code-latest","name":"Auto","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"hunyuan-t1":{"id":"hunyuan-t1","name":"Hunyuan-T1","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-03-08","last_updated":"2026-03-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"crossmodel":{"id":"crossmodel","env":["CROSSMODEL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.crossmodel.ai/v1","name":"CrossModel","doc":"https://www.crossmodel.ai/docs","models":{"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.2,"output":4.4,"cache_read":0.3,"cache_write":1.2},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.47,"output":2.16,"cache_read":0.1,"cache_write":0.47,"tiers":[{"input":0.62,"output":2.47,"cache_read":0.13,"cache_write":0.62,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1,"output":3.8,"cache_read":0.2,"cache_write":1,"tiers":[{"input":1.2,"output":4.4,"cache_read":0.3,"cache_write":1.2,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":3,"cache_read":0.16,"cache_write":0.6,"tiers":[{"input":0.8,"output":3.4,"cache_read":0.2,"cache_write":0.8,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.2,"output":4.4,"cache_read":0.3,"cache_write":1.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.9,"output":3.7,"cache_read":0.18,"cache_write":0.9,"tiers":[{"input":1.1,"output":4.3,"cache_read":0.27,"cache_write":1.1,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0.15},"canonical_model_id":"zhipuai/glm-5.3-flash"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.47,"output":0.94,"cache_read":0.005,"cache_write":0.47},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.16,"output":0.32,"cache_read":0.004,"cache_write":0.16},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0.96,"output":2.88,"cache_read":0.048,"cache_write":0.96},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":131072},"cost":{"input":0.16,"output":0.64,"cache_read":0.04,"cache_write":0.16},"canonical_model_id":"tencent/hy3"},"gemini/gemini-4-argon":{"id":"gemini/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2},"canonical_model_id":"google/gemini-4-argon"},"gemini/gemini-3-flash-preview":{"id":"gemini/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.5},"canonical_model_id":"google/gemini-3-flash-preview"},"gemini/gemini-3.5-flash":{"id":"gemini/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gemini/gemini-3.8-flash":{"id":"gemini/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"gemini/gemini-3.1-pro-preview":{"id":"gemini/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gemini/gemini-3.7-flash":{"id":"gemini/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"gemini/gemini-2.5-flash":{"id":"gemini/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"google/gemini-2.5-flash"},"gemini/gemini-2.5-pro":{"id":"gemini/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":1.25,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5}},"canonical_model_id":"google/gemini-2.5-pro"},"gemini/gemini-3.6-flash":{"id":"gemini/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"gemini/gemini-3.5-flash-lite":{"id":"gemini/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gemini/gemini-2.5-flash-lite":{"id":"gemini/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.1},"canonical_model_id":"google/gemini-2.5-flash-lite"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.35,"output":4.05,"cache_read":0.045,"cache_write":1.35},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.3},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5,"cache_write":5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":10}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02,"cache_write":0.2},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075,"cache_write":0.15},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":512000},"cost":{"input":0.33,"output":1.32,"cache_read":0.066,"cache_write":0.33,"tiers":[{"input":0.66,"output":2.63,"cache_read":0.132,"cache_write":0.66,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.66,"output":2.63,"cache_read":0.132,"cache_write":0.66}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.33,"output":1.32,"cache_read":0.066,"cache_write":0.42},"canonical_model_id":"minimax/MiniMax-M2.7"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.32,"output":1.25,"cache_read":0.032,"cache_write":0.4,"tiers":[{"input":0.96,"output":3.75,"cache_read":0.096,"cache_write":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.96,"output":3.75,"cache_read":0.096,"cache_write":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.13,"output":0.43,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.13,"output":0.43,"cache_read":0.016,"cache_write":0.13},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.32,"output":1.88,"cache_read":0.032,"cache_write":0.4,"tiers":[{"input":1.25,"output":7.5,"cache_read":0.124,"cache_write":1.57,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.25,"output":7.5,"cache_read":0.124,"cache_write":1.57}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.19,"output":1.13,"cache_read":0.019,"cache_write":0.24,"tiers":[{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.94,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.94}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.88,"output":5.63,"cache_read":0.375,"cache_write":2.35},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.88,"output":5.63,"cache_read":0.23,"cache_write":2.35},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.04,"output":0.13,"cache_read":0.01,"cache_write":0.04,"tiers":[{"input":0.1,"output":0.37,"cache_read":0.02,"cache_write":0.12,"tier":{"type":"context","size":32000}},{"input":0.19,"output":0.74,"cache_read":0.04,"cache_write":0.24,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"cache_write":1.25,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"cache_write":1,"tiers":[{"input":2,"output":4,"cache_read":0.4,"cache_write":2,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4,"cache_write":2}},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":0.6,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6,"cache_write":4}},"canonical_model_id":"xai/grok-4.5"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":1,"output":4.16,"cache_read":0.18,"cache_write":1},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":1,"output":4.16,"cache_read":0.18,"cache_write":1},"canonical_model_id":"moonshotai/kimi-k2.6"}}},"pareto":{"id":"pareto","env":["PARETO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.paretoinference.com/v1","name":"Pareto Inference","doc":"https://docs.paretoinference.com/","models":{"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.06,"output":0.2,"cache_read":0.012},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"melious":{"id":"melious","env":["MELIOUS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.melious.ai/v1","name":"Melious","doc":"https://melious.ai/docs/reference/models","models":{"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.728975,"output":3.644875,"cache_read":0.190655},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.23184,"output":1.1592,"cache_read":0.011592},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.33675,"output":1.347,"cache_read":0.033675},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"status":"deprecated","cost":{"input":1.50696,"output":4.6368,"cache_read":0.370944},"canonical_model_id":"zhipuai/glm-5.1"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"status":"deprecated","cost":{"input":0.5796,"output":2.95596,"cache_read":0.139104},"canonical_model_id":"moonshotai/kimi-k2.5"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3.196275,"output":15.981375,"cache_read":0.841125},"canonical_model_id":"moonshotai/kimi-k3"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.78505,"output":3.3645,"cache_read":0.280375},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.11592,"output":0.2898,"cache_read":0.023184},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131072},"status":"deprecated","cost":{"input":1.10124,"output":3.36168,"cache_read":0.266616},"canonical_model_id":"zhipuai/glm-5"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1592,"output":3.4776,"cache_read":0.23184},"canonical_model_id":"zhipuai/glm-5.3"},"deepseek-r1-0528":{"id":"deepseek-r1-0528","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":32768},"status":"deprecated","cost":{"input":0.69552,"output":2.78208,"cache_read":0.185472},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":64000},"status":"deprecated","cost":{"input":0.34776,"output":0.5796,"cache_read":0.092736},"canonical_model_id":"deepseek/deepseek-v3.2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11592,"output":0.46368,"cache_read":0.023184},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"openrouter":{"id":"openrouter","env":["OPENROUTER_API_KEY"],"npm":"@openrouter/ai-sdk-provider","api":"https://openrouter.ai/api/v1","name":"OpenRouter","doc":"https://openrouter.ai/models","models":{"mistralai/mistral-large-2407":{"id":"mistralai/mistral-large-2407","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-2407"},"mistralai/mistral-large-4-0":{"id":"mistralai/mistral-large-4-0","name":"Mistral Large 4","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 25.08","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":204800},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"mistralai/voxtral-small-24b-2507":{"id":"mistralai/voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":26214},"cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Mistral Small 3.2 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.075,"output":0.2},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Mistral Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":102400},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":102400},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-8b-2512"},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":26214},"cost":{"input":0.2,"output":0.6,"cache_read":0.02},"canonical_model_id":"mistral/mistral-saba"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Mistral Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral Nemo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.029,"output":0.03},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistralai/mistral-medium-3-5":{"id":"mistralai/mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3b-2512"},"recursive/flint-v1.0:free":{"id":"recursive/flint-v1.0:free","name":"Flint (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"flint","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember-1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":117964},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.0341,"output":7,"cache_read":0.033},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.54,"output":1.98,"cache_read":0.099},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.3-prime":{"id":"z-ai/glm-5.3-prime","name":"GLM 5.3 Prime","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.56}},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"status":"deprecated","cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM 5.3 FlashX","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.09}},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.079,"output":1.4,"cache_read":0.078},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/lyria-3-pro-preview":{"id":"google/lyria-3-pro-preview","name":"Lyria 3 Pro Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-pro-preview"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.08,"output":0.35,"cache_read":0.04},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"reasoning":3,"cache_read":0.05,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"reasoning":9,"cache_read":0.15,"cache_write":0.083333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.825,"output":4.125,"reasoning":4.125,"cache_read":0.0825,"cache_write":0.041667,"input_audio":0.825},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-2.5-pro-preview":{"id":"google/gemini-2.5-pro-preview","name":"Gemini 2.5 Pro Preview 06-05","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["pdf","image","text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"reasoning":10,"cache_read":0.125,"cache_write":0.375,"input_audio":1.25,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro-preview"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/lyria-3-clip-preview":{"id":"google/lyria-3-clip-preview","name":"Lyria 3 Clip Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-clip-preview"},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemma-4-31b-it:free":{"id":"google/gemma-4-31b-it:free","name":"Gemma 4 31B (free)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.04,"output":0.05,"cache_read":0.02},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"reasoning":10,"cache_read":0.125,"cache_write":0.375,"input_audio":1.25,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemma-4-26b-a4b-it:free":{"id":"google/gemma-4-26b-a4b-it:free","name":"Gemma 4 26B A4B (free)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375,"input_audio":2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.0675,"output":0.225,"cache_read":0.0375},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"reasoning":2.5,"cache_read":0.03,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.083333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.1,"output":0.4,"reasoning":0.4,"cache_read":0.01,"cache_write":0.083333,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"google/gemma-2-27b-it":{"id":"google/gemma-2-27b-it","name":"Gemma 2 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-07-13","last_updated":"2024-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0.65,"output":0.65},"canonical_model_id":"google/gemma-2-27b-it"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["image","text"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"nvidia/nemotron-3.5-super-vl:free":{"id":"nvidia/nemotron-3.5-super-vl:free","name":"Nemotron 3.5 Super VL (free)","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-3-ultra-550b-a55b:free":{"id":"nvidia/nemotron-3-ultra-550b-a55b:free","name":"Nemotron 3 Ultra (free)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3.5-content-safety:free":{"id":"nvidia/nemotron-3.5-content-safety:free","name":"Nemotron 3.5 Content Safety (free)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3.5-lightning:free":{"id":"nvidia/nemotron-3.5-lightning:free","name":"Nemotron 3.5 Lightning (free)","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free":{"id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free","name":"Nemotron 3 Nano Omni (free)","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.3,"output":2.4,"cache_read":0.15},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":182520},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-super-120b-a12b:free":{"id":"nvidia/nemotron-3-super-120b-a12b:free","name":"Nemotron 3 Super (free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nemotron 3.5 Lightning 30B A3B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.07,"output":0.2,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"Nemotron 3.5 Super VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":235929},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"inception/mercury-2":{"id":"inception/mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.04,"output":0.15,"cache_read":0.004},"canonical_model_id":"inception/mercury-2.5"},"inception/mercury-3":{"id":"inception/mercury-3","name":"Mercury 3","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01}},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.119,"output":0.238,"cache_read":0.00255},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.3045,"output":0.609,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"xingchen-agi/xing4.0-29b-a4b:free":{"id":"xingchen-agi/xing4.0-29b-a4b:free","name":"Xing4.0 29B A4B (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"agi","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"elorian/glimpse":{"id":"elorian/glimpse","name":"Glimpse","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":256000},"cost":{"input":0.6,"output":3,"cache_read":0.06}},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":256000,"output":230400},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1000000,"output":64000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"meituan/longcat-2.0":{"id":"meituan/longcat-2.0","name":"LongCat-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048756,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.0"},"perceptron/perceptron-mk1":{"id":"perceptron/perceptron-mk1","name":"Perceptron Mk1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"perceptron/perceptron-mk1"},"perceptron/perceptron-mk1.5":{"id":"perceptron/perceptron-mk1.5","name":"Perceptron Mk1.5","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":36864,"output":8192},"cost":{"input":0.15,"output":1.5}},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling:free":{"id":"thinkingmachines/inkling:free","name":"Inkling (free)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/inkling-small:free":{"id":"thinkingmachines/inkling-small:free","name":"Inkling Small (free)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":471859},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"mancer/weaver":{"id":"mancer/weaver","name":"Weaver (alpha)","description":"General-purpose chat model for instruction following, writing, and analysis","family":"alpha","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-08-02","last_updated":"2023-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":6000},"cost":{"input":0.4,"output":0.75},"canonical_model_id":"mancer/weaver"},"unbiased/pareto-26.9":{"id":"unbiased/pareto-26.9","name":"Pareto 26.9","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25}},"unbiased/pareto-26.10":{"id":"unbiased/pareto-26.10","name":"Pareto 26.10","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"relace/relace-apply-3":{"id":"relace/relace-apply-3","name":"Relace Apply 3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.85,"output":1.25},"canonical_model_id":"relace/relace-apply-3"},"relace/relace-search":{"id":"relace/relace-search","name":"Relace Search","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":1,"output":3},"canonical_model_id":"relace/relace-search"},"~x-ai/grok-latest":{"id":"~x-ai/grok-latest","name":"Grok Latest","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}}},"~moonshotai/kimi-latest":{"id":"~moonshotai/kimi-latest","name":"Kimi Latest","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.19,"output":8.94,"cache_read":0.19}},"reflection/beam-1.0-501b-a23b":{"id":"reflection/beam-1.0-501b-a23b","name":"Beam 1.0 501B A23B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.25,"output":1,"cache_read":0.05},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"tencent/hunyuan-a13b-instruct":{"id":"tencent/hunyuan-a13b-instruct","name":"Hunyuan A13B Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.14,"output":0.57}},"tencent/hy-mt2-7b":{"id":"tencent/hy-mt2-7b","name":"Hy-MT2-7B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-7b"},"tencent/hy3-preview":{"id":"tencent/hy3-preview","name":"Hy3 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.18,"output":0.6,"cache_read":0.06},"canonical_model_id":"tencent/hy3-preview"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy-mt2-30b-a3b":{"id":"tencent/hy-mt2-30b-a3b","name":"Hy-MT2-30B-A3B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-30b-a3b"},"tencent/hy-mt2-1.8b":{"id":"tencent/hy-mt2-1.8b","name":"Hy-MT2-1.8B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.044,"output":0.177},"canonical_model_id":"tencent/hy-mt2-1.8b"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":80000},"cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-4-maverick":{"id":"meta-llama/llama-4-maverick","name":"Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":16384},"cost":{"input":0.1875,"output":0.6525,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama-3.1-8B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.05,"output":0.08,"cache_read":0.025},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-3.2-1b-instruct":{"id":"meta-llama/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":54000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"meta-llama/llama-4-scout":{"id":"meta-llama/llama-4-scout","name":"Llama 4 Scout","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1310720,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.1-70b-instruct":{"id":"meta-llama/llama-3.1-70b-instruct","name":"Llama-3.1-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"meta-llama/llama-3.2-3b-instruct":{"id":"meta-llama/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":72000},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"meta-llama/llama-guard-4-12b":{"id":"meta-llama/llama-guard-4-12b","name":"Llama Guard 4 12B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":0.41}},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.02,"output":1,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v3.1-terminus":{"id":"deepseek/deepseek-v3.1-terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":1}},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.31755,"output":0.6351,"cache_read":0.026463},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-chat-v3.1":{"id":"deepseek/deepseek-chat-v3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-chat-v3.1"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.04444,"output":0.13332,"cache_read":0.001414},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-chat-v3-0324":{"id":"deepseek/deepseek-chat-v3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":115200},"cost":{"input":0.29,"output":1.14,"cache_read":0.11},"canonical_model_id":"deepseek/deepseek-chat-v3-0324"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16000},"cost":{"input":0.2574,"output":1.0287},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.091,"output":0.182,"cache_read":0.0182},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.00686},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"nousresearch/hermes-3-llama-3.1-70b":{"id":"nousresearch/hermes-3-llama-3.1-70b","name":"Hermes 3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":0.7},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"nousresearch/hermes-4-405b":{"id":"nousresearch/hermes-4-405b","name":"Hermes 4 405B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":1,"output":3},"canonical_model_id":"nousresearch/hermes-4-405b"},"nousresearch/hermes-3-llama-3.1-405b":{"id":"nousresearch/hermes-3-llama-3.1-405b","name":"Hermes 3 405B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":1,"output":1},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"rekaai/reka-flash-3":{"id":"rekaai/reka-flash-3","name":"Reka Flash 3","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"reka","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":58982},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"rekaai/reka-flash-3"},"rekaai/reka-edge":{"id":"rekaai/reka-edge","name":"Reka Edge","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":14745},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"rekaai/reka-edge"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 Instruct 0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.239,"output":10.8,"cache_read":0.21},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.65,"output":3.41,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-08-21","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-wonder":{"id":"meta/muse-wonder","name":"Muse Wonder","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":4,"output":20,"cache_read":0.2}},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"openrouter/bodybuilder":{"id":"openrouter/bodybuilder","name":"Body Builder (beta)","description":"Preview model for early access evaluation, prototyping, and compatibility testing","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-12-05","last_updated":"2025-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000}},"openrouter/free":{"id":"openrouter/free","name":"Free Models Router","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-01","last_updated":"2026-02-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":8000},"cost":{"input":0,"output":0}},"openrouter/fusion":{"id":"openrouter/fusion","name":"Fusion","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"openrouter/auto":{"id":"openrouter/auto","name":"Auto Router","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2023-11-08","last_updated":"2023-11-08","modalities":{"input":["text","image","audio","pdf","video"],"output":["text","image"]},"open_weights":false,"limit":{"context":2000000,"output":2000000}},"openrouter/pareto-code":{"id":"openrouter/pareto-code","name":"Pareto Code Router","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":200000}},"~anthropic/claude-fable-latest":{"id":"~anthropic/claude-fable-latest","name":"Claude Fable Latest","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5}},"~anthropic/claude-sonnet-latest":{"id":"~anthropic/claude-sonnet-latest","name":"Claude Sonnet Latest","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"~anthropic/claude-haiku-latest":{"id":"~anthropic/claude-haiku-latest","name":"Claude Haiku Latest","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]}},"~anthropic/claude-opus-latest":{"id":"~anthropic/claude-opus-latest","name":"Claude Opus Latest","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5}},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-27","last_updated":"2025-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127072,"output":114364},"cost":{"input":1,"output":1}},"perplexity/sonar-reasoning-pro":{"id":"perplexity/sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded reasoning model for multi-step research and cited answers","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8}},"perplexity/sonar-deep-research":{"id":"perplexity/sonar-deep-research","name":"Sonar Deep Research","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar-deep-research","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8,"reasoning":3}},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Sonar Pro","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"perplexity/sonar-pro-search":{"id":"perplexity/sonar-pro-search","name":"Sonar Pro Search","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-30","last_updated":"2025-10-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":117964},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-mini4":{"id":"upstage/solar-mini4","name":"Solar Mini 4","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.005},"canonical_model_id":"upstage/solar-mini4"},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.15,"output":0.6,"cache_read":0.03},"canonical_model_id":"upstage/solar-pro4"},"dots-studio/dots-3-note-preview:free":{"id":"dots-studio/dots-3-note-preview:free","name":"Dots3-Note Preview (free)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":460800},"status":"deprecated","cost":{"input":0,"output":0}},"liquid/lfm-2.5-24b-a2b":{"id":"liquid/lfm-2.5-24b-a2b","name":"LFM2.5-24B-A2B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"liquid/lfm-2.5-2.6b:free":{"id":"liquid/lfm-2.5-2.6b:free","name":"LFM2.5-2.6B (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"cost":{"input":0,"output":0}},"cognitivecomputations/dolphin-mistral-24b-venice-edition":{"id":"cognitivecomputations/dolphin-mistral-24b-venice-edition","name":"Uncensored","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2,"output":0.9},"canonical_model_id":"cognitivecomputations/dolphin-mistral-24b-venice-edition"},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"ibm-granite/granite-4.0-h-micro":{"id":"ibm-granite/granite-4.0-h-micro","name":"Granite 4.0 Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"granite","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":117900},"cost":{"input":0.017,"output":0.112}},"aion-labs/aion-3.0-mini":{"id":"aion-labs/aion-3.0-mini","name":"Aion-3.0-Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18},"canonical_model_id":"aion-labs/aion-3.0-mini"},"aion-labs/aion-3.5-mini":{"id":"aion-labs/aion-3.5-mini","name":"Aion 3.5 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18}},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"Aion-RP 1.0 (8B)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.8,"output":1.6},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"aion-labs/aion-3.0":{"id":"aion-labs/aion-3.0","name":"Aion-3.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75},"canonical_model_id":"aion-labs/aion-3.0"},"aion-labs/aion-2.0":{"id":"aion-labs/aion-2.0","name":"Aion-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-21","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.8,"output":1.6,"cache_read":0.2},"canonical_model_id":"aion-labs/aion-2.0"},"aion-labs/aion-3.5":{"id":"aion-labs/aion-3.5","name":"Aion 3.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75}},"moondream/moondream3.1-9b-a2b":{"id":"moondream/moondream3.1-9b-a2b","name":"Moondream 3.1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"moondream/moondream3-preview":{"id":"moondream/moondream3-preview","name":"Moondream 3 Preview","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"~deepseek/deepseek-flash-latest":{"id":"~deepseek/deepseek-flash-latest","name":"DeepSeek Flash Latest","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.02,"output":1,"cache_read":0.015}},"~deepseek/deepseek-pro-latest":{"id":"~deepseek/deepseek-pro-latest","name":"DeepSeek Pro Latest","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":393216},"cost":{"input":0.1188,"output":8,"cache_read":0.12}},"~deepseek/deepseek-v4-flash-latest":{"id":"~deepseek/deepseek-v4-flash-latest","name":"DeepSeek V4 Flash Latest","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-01","last_updated":"2026-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.0422,"output":0.1694,"cache_read":0.028}},"amazon/nova-micro-v1":{"id":"amazon/nova-micro-v1","name":"Nova Micro 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":5120},"cost":{"input":0.035,"output":0.14},"canonical_model_id":"amazon/nova-micro-v1"},"amazon/nova-premier-v1":{"id":"amazon/nova-premier-v1","name":"Nova Premier 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-30","last_updated":"2025-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"status":"deprecated","cost":{"input":2.5,"output":12.5,"cache_read":0.625},"canonical_model_id":"amazon/nova-premier"},"amazon/nova-2-lite-v1":{"id":"amazon/nova-2-lite-v1","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"amazon/nova-2-lite-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.8,"output":3.2},"canonical_model_id":"amazon/nova-pro-v1"},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"amazon/nova-lite-v1"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.021,"output":0.063,"cache_read":0.0042},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.075,"output":0.225,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.1-flash"},"~google/gemini-pro-latest":{"id":"~google/gemini-pro-latest","name":"Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["audio","pdf","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}}},"~google/gemini-flash-latest":{"id":"~google/gemini-flash-latest","name":"Gemini Flash Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667,"input_audio":0.75}},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"output":5529},"cost":{"input":0.35,"output":0.65},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"~unbiased/pareto-latest":{"id":"~unbiased/pareto-latest","name":"Pareto Latest","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"openai/gpt-6-astra-pro":{"id":"openai/gpt-6-astra-pro","name":"GPT-6 Astra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}}},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-4o-mini-2024-07-18":{"id":"openai/gpt-4o-mini-2024-07-18","name":"GPT-4o-mini (2024-07-18)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":150,"output":600},"canonical_model_id":"openai/o1-pro"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5-image-mini":{"id":"openai/gpt-5-image-mini","type":"image","name":"GPT-5 Image Mini","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["pdf","image","text"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":2,"cache_read":0.25},"canonical_model_id":"openai/gpt-5-image-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5-image":{"id":"openai/gpt-5-image","type":"image","name":"GPT-5 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-10-01","release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":10,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-5-image"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.018,"output":0.09,"cache_read":0.009},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-audio-mini":{"id":"openai/gpt-audio-mini","type":"chat","name":"GPT-Audio Mini","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":2.4,"input_audio":0.6,"output_audio":2.4},"canonical_model_id":"openai/gpt-audio-mini"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6.1-sol-pro":{"id":"openai/gpt-6.1-sol-pro","name":"GPT-6.1 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}}},"openai/gpt-audio":{"id":"openai/gpt-audio","type":"chat","name":"GPT-Audio","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-audio"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-6-luna-pro":{"id":"openai/gpt-6-luna-pro","name":"GPT-6 Luna Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}}},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/o4-mini-high":{"id":"openai/o4-mini-high","name":"o4 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-06-30","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275}},"openai/gpt-chat-latest":{"id":"openai/gpt-chat-latest","name":"GPT Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5}},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.4-image-2":{"id":"openai/gpt-5.4-image-2","type":"image","name":"GPT-5.4 Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":8,"output":15,"cache_read":2},"canonical_model_id":"openai/gpt-5.4-image-2"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-3.5-turbo-16k":{"id":"openai/gpt-3.5-turbo-16k","name":"GPT-3.5 Turbo 16k","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-30","release_date":"2023-08-28","last_updated":"2023-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":3,"output":4}},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":4096},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-6-sol-pro":{"id":"openai/gpt-6-sol-pro","name":"GPT-6 Sol Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}}},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2-chat":{"id":"openai/gpt-5.2-chat","name":"GPT-5.2 Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-3.5-turbo-instruct":{"id":"openai/gpt-3.5-turbo-instruct","name":"GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2021-09-30","release_date":"2023-09-28","last_updated":"2023-09-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1.5,"output":2}},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","pdf","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-3.5-turbo-0613":{"id":"openai/gpt-3.5-turbo-0613","name":"GPT-3.5 Turbo (older v0613)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-30","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1,"output":2}},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"o3 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-10-31","release_date":"2025-02-12","last_updated":"2025-02-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"prism-ml/ternary-bonsai-2-27b":{"id":"prism-ml/ternary-bonsai-2-27b","name":"Ternary Bonsai 2 27B","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.5,"cache_read":0.0375}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.05},"canonical_model_id":"inference-net/schematron-v2-small"},"microsoft/phi-4":{"id":"microsoft/phi-4","name":"Phi 4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-01-10","last_updated":"2025-01-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":14745},"cost":{"input":0.07,"output":0.14}},"poolside/laguna-s-2.1:free":{"id":"poolside/laguna-s-2.1:free","name":"Laguna S 2.1 (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-xs-2.1:free":{"id":"poolside/laguna-xs-2.1:free","name":"Laguna XS 2.1 (free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.06,"output":0.12,"cache_read":0.03},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.09,"output":0.18,"cache_read":0.009},"canonical_model_id":"poolside/laguna-s-2.1"},"~openai/gpt-sol-latest":{"id":"~openai/gpt-sol-latest","name":"GPT Sol Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}}},"~openai/gpt-terra-latest":{"id":"~openai/gpt-terra-latest","name":"GPT Terra Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}}},"~openai/gpt-luna-latest":{"id":"~openai/gpt-luna-latest","name":"GPT Luna Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}}},"~openai/gpt-mini-latest":{"id":"~openai/gpt-mini-latest","name":"GPT Mini Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075}},"~openai/gpt-astra-latest":{"id":"~openai/gpt-astra-latest","name":"GPT Astra Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}}},"thedrummer/skyfall-36b-v2":{"id":"thedrummer/skyfall-36b-v2","name":"Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.55,"output":0.8,"cache_read":0.25},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"thedrummer/cydonia-24b-v4.1":{"id":"thedrummer/cydonia-24b-v4.1","name":"Cydonia 24B V4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.3,"output":0.5,"cache_read":0.15},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"thedrummer/unslopnemo-12b":{"id":"thedrummer/unslopnemo-12b","name":"UnslopNemo 12B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-11-08","last_updated":"2024-11-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":819200},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"thedrummer/unslopnemo-12b"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax L2 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":3686},"cost":{"input":0.08,"output":0.11},"canonical_model_id":"gryphe/mythomax-l2-13b"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/sakana-namazu":{"id":"sakana/sakana-namazu","name":"Sakana Namazu","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"minimax/minimax-01":{"id":"minimax/minimax-01","name":"MiniMax-01","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-03-31","release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000192,"output":40000},"cost":{"input":0.2,"output":1.1}},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m1":{"id":"minimax/minimax-m1","name":"MiniMax M1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":40000},"cost":{"input":0.55,"output":2.2}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2-her":{"id":"minimax/minimax-m2-her","name":"MiniMax-M2 Her","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":2048},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2-Her"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.255,"output":1.02},"canonical_model_id":"minimax/MiniMax-M2"},"sao10k/l3-lunaris-8b":{"id":"sao10k/l3-lunaris-8b","name":"Llama 3 8B Lunaris","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-13","last_updated":"2024-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":7372},"cost":{"input":0.04,"output":0.05},"canonical_model_id":"sao10k/l3-lunaris-8b"},"sao10k/l3.1-euryale-70b":{"id":"sao10k/l3.1-euryale-70b","name":"Llama 3.1 Euryale 70B v2.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.85,"output":0.85},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"status":"deprecated","cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.21,"output":1.9,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.32,"output":1.28,"cache_read":0.064,"cache_write":0.4,"tiers":[{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3-coder":{"id":"qwen/qwen3-coder","name":"Qwen3 Coder 480B A35B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":1,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen/qwen3.5-plus-02-15":{"id":"qwen/qwen3.5-plus-02-15","name":"Qwen3.5 Plus 2026-02-15","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.26,"output":1.56,"tiers":[{"input":0.325,"output":1.95,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.325,"output":1.95}}},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.26,"output":0.78,"cache_read":0.052,"cache_write":0.325,"tiers":[{"input":0.78,"output":2.34,"cache_read":0.156,"cache_write":0.975,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.78,"output":2.34,"cache_read":0.156,"cache_write":0.975}},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3-coder-30b-a3b-instruct":{"id":"qwen/qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"qwen/qwen3.5-plus-20260420":{"id":"qwen/qwen3.5-plus-20260420","name":"Qwen3.5 Plus 2026-04-20","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.8,"cache_write":0.375,"tiers":[{"input":0.375,"output":2.25,"cache_write":0.46875,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.375,"output":2.25,"cache_write":0.46875}}},"qwen/qwen3.8-max-prime":{"id":"qwen/qwen3.8-max-prime","name":"Qwen 3.8 Max Prime","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"qwen/qwen2.5-vl-72b-instruct":{"id":"qwen/qwen2.5-vl-72b-instruct","name":"Qwen2.5-VL 72B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":115200},"cost":{"input":0.8,"output":1,"cache_read":0.4},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"qwen/qwen-2.5-72b-instruct":{"id":"qwen/qwen-2.5-72b-instruct","name":"Qwen2.5 72B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.36,"output":0.4}},"qwen/qwen3.5-flash-02-23":{"id":"qwen/qwen3.5-flash-02-23","name":"Qwen3.5-Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-25","last_updated":"2026-02-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.065,"output":0.26}},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B-A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.09,"output":0.3},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":909000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.12,"output":0.8,"cache_read":0.07},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"cache_write":0.40625,"tiers":[{"input":1.3,"output":3.9,"cache_write":1.625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.3,"output":3.9,"cache_write":1.625}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":3.6,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-14b":{"id":"qwen/qwen3-14b","name":"Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.24}},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1875,"output":1.125,"cache_write":0.234375,"tiers":[{"input":0.75,"output":3,"cache_write":0.9375,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.75,"output":3,"cache_write":0.9375}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B-A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.45,"output":3.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen/qwen3-vl-30b-a3b-thinking":{"id":"qwen/qwen3-vl-30b-a3b-thinking","name":"Qwen3 VL 30B A3B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.29,"output":1}},"qwen/qwen-2.5-7b-instruct":{"id":"qwen/qwen-2.5-7b-instruct","name":"Qwen2.5 7B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-16","last_updated":"2024-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.1,"output":0.2}},"qwen/qwen3-vl-8b-instruct":{"id":"qwen/qwen3-vl-8b-instruct","name":"Qwen3 VL 8B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.25,"output":0.75,"cache_read":0.12}},"qwen/qwen3-235b-a22b-2507":{"id":"qwen/qwen3-235b-a22b-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.1,"output":0.95,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.27,"output":2.16,"cache_read":0.27},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen-2.5-coder-32b-instruct":{"id":"qwen/qwen-2.5-coder-32b-instruct","name":"Qwen2.5-Coder-32B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.195,"output":0.975,"cache_read":0.039,"cache_write":0.24375,"tiers":[{"input":0.325,"output":1.625,"cache_read":0.065,"cache_write":0.40625,"tier":{"type":"context","size":32000}},{"input":0.52,"output":2.6,"cache_read":0.104,"cache_write":0.65,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":235929},"cost":{"input":0.2,"output":2.5,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.20-multi-agent":{"id":"x-ai/grok-4.20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":230400},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"bytedance-seed/seed-1.6":{"id":"bytedance-seed/seed-1.6","name":"Seed 1.6","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.25,"output":2,"tiers":[{"input":0.5,"output":4,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"Seed 2.0 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.25,"output":2,"tiers":[{"input":0.5,"output":4,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-mini":{"id":"bytedance-seed/seed-2.0-mini","name":"Seed 2.0 Mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.4,"tiers":[{"input":0.2,"output":0.8,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance-seed/seed-1.6-flash":{"id":"bytedance-seed/seed-1.6-flash","name":"Seed 1.6 Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.075,"output":0.3,"tiers":[{"input":0.1,"output":0.8,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.5,"output":3,"tiers":[{"input":1,"output":6,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"bytedance-seed/seed-2-1-turbo":{"id":"bytedance-seed/seed-2-1-turbo","name":"Seed 2.1 Turbo","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":2.5},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":2.5,"output":5},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"scaledown/extract":{"id":"scaledown/extract","name":"Extract","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":0.04,"output":0}},"scaledown/summarize":{"id":"scaledown/summarize","name":"Summarize","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":4000},"cost":{"input":0.04,"output":0}},"~z-ai/glm-flash-latest":{"id":"~z-ai/glm-flash-latest","name":"GLM Flash Latest","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.0393,"output":0.1028,"cache_read":0.03}},"~z-ai/glm-latest":{"id":"~z-ai/glm-latest","name":"GLM Latest","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-19","last_updated":"2026-08-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.075,"output":4.295,"cache_read":0.075}},"bytedance/ui-tars-1.5-7b":{"id":"bytedance/ui-tars-1.5-7b","name":"UI-TARS 7B ","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"cost":{"input":0.1,"output":0.2,"cache_read":0.1}},"writer/palmyra-x5":{"id":"writer/palmyra-x5","name":"Palmyra X5","description":"General-purpose chat model for instruction following, writing, and analysis","family":"palmyra","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"output":8192},"cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"cohere/command-a-plus":{"id":"cohere/command-a-plus","name":"Command A+","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04-01","release_date":"2026-05-20","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":64000},"cost":{"input":0.3,"output":1.5,"cache_read":0.15},"canonical_model_id":"cohere/command-a-plus-05-2026"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a"},"cohere/north-mini-code:free":{"id":"cohere/north-mini-code:free","name":"North Mini Code (free)","description":"Cohere coding model for practical software engineering and agentic edits","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"nex-agi/nex-n2.5-mini":{"id":"nex-agi/nex-n2.5-mini","name":"Nex-N2.5-Mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.025,"output":0.1,"cache_read":0.0025}},"nex-agi/nex-n2.5-pro":{"id":"nex-agi/nex-n2.5-pro","name":"Nex-N2.5-Pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.075,"output":0.25,"cache_read":0.015}},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"Morph V3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.9,"output":1.9},"canonical_model_id":"morph/morph-v3-large"},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"Morph V3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.8,"output":1.2},"canonical_model_id":"morph/morph-v3-fast"}}},"google-vertex-anthropic":{"id":"google-vertex-anthropic","env":["GOOGLE_VERTEX_PROJECT","GOOGLE_VERTEX_LOCATION","GOOGLE_APPLICATION_CREDENTIALS"],"npm":"@ai-sdk/google-vertex/anthropic","name":"Vertex (Anthropic)","doc":"https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude","models":{"claude-haiku-4-5@20251001":{"id":"claude-haiku-4-5@20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"claude-fable-5@default":{"id":"claude-fable-5@default","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-4-7@default":{"id":"claude-opus-4-7@default","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"claude-sonnet-5-5@default":{"id":"claude-sonnet-5-5@default","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-sonnet-4-5@20250929":{"id":"claude-sonnet-4-5@20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-opus-4-5@20251101":{"id":"claude-opus-4-5@20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"claude-haiku-5-5@default":{"id":"claude-haiku-5-5@default","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-opus-5@default":{"id":"claude-opus-5@default","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus-4-6@default":{"id":"claude-opus-4-6@default","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5-5@default":{"id":"claude-opus-5-5@default","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-1@20250805":{"id":"claude-opus-4-1@20250805","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"claude-opus-4-8@default":{"id":"claude-opus-4-8@default","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-sonnet-5@default":{"id":"claude-sonnet-5@default","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-sonnet-4@20250514":{"id":"claude-sonnet-4@20250514","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"claude-fable-5-1@default":{"id":"claude-fable-5-1@default","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4@20250514":{"id":"claude-opus-4@20250514","name":"Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"claude-sonnet-4-6@default":{"id":"claude-sonnet-4-6@default","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"}}},"baseten":{"id":"baseten","env":["BASETEN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.baseten.co/v1","name":"Baseten","doc":"https://docs.baseten.co/inference/model-apis/overview","models":{"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B","name":"Nemotron Ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":202800},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-ai/DeepSeek-V4.1-Flash-1017-Fast":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017-Fast","name":"DeepSeek V4.1 Flash 1017 Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.3,"output":1.2,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4.1-Flash-Fast":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-Fast","name":"DeepSeek V4.1 Flash Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.32,"output":3.96,"cache_read":0.132},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"OpenAI GPT 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128072,"output":128072},"cost":{"input":0.1,"output":0.5,"cache_read":0.1},"canonical_model_id":"openai/gpt-oss-120b"},"zai-org/GLM-5.2-Fast":{"id":"zai-org/GLM-5.2-Fast","name":"GLM 5.2 Fast","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5.3-Fast":{"id":"zai-org/GLM-5.3-Fast","name":"GLM 5.3 Fast","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM 5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"}}},"minimax-coding-plan":{"id":"minimax-coding-plan","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.io/anthropic/v1","name":"MiniMax Token Plan (minimax.io)","doc":"https://platform.minimax.io/docs/token-plan/intro","models":{"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash-Preview":{"id":"MiniMax-M3.1-Flash-Preview","name":"MiniMax-M3.1-Flash-Preview","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-27","last_updated":"2026-09-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash-Preview"},"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}}}},"the-grid-ai":{"id":"the-grid-ai","env":["THEGRID_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.thegrid.ai/v1","name":"The Grid AI","doc":"https://thegrid.ai/docs","models":{"agent-prime":{"id":"agent-prime","name":"Agent Prime","description":"Reliable models for dependable agentic applications, multi-step tool use, and reasoning workflows. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"input":120000,"output":131072},"status":"beta"},"text-prime":{"id":"text-prime","name":"Text Prime","description":"Reliable models for everyday text generation, editing, and analysis across diverse workflows. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-26","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"input":120000,"output":131072}},"kimi-latest":{"id":"kimi-latest","name":"Kimi Latest","description":"Lab latest market for Kimi supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":922000,"output":943718}},"code-prime":{"id":"code-prime","name":"Code Prime","description":"Reliable models for everyday software tasks, code completion, review, and standard debugging. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"input":120000,"output":131072},"status":"beta"},"agent-max":{"id":"agent-max","name":"Agent Max","description":"Frontier models for autonomous research, deep multi-step tool chains, and complex long-horizon tasks. Any model that meets the contract spec can serve your request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-05-04","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"status":"beta"},"gpt-sol-latest":{"id":"gpt-sol-latest","name":"GPT Sol Latest","description":"Lab latest market for GPT frontier supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000}},"minimax-latest":{"id":"minimax-latest","name":"MiniMax Latest","description":"Lab latest market for MiniMax supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":512000}},"agent-standard":{"id":"agent-standard","name":"Agent Standard","description":"Price-optimized models for fast tool calls, simple agent loops, high-throughput automation, and orchestration. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":120000,"output":65536},"status":"beta"},"code-standard":{"id":"code-standard","name":"Code Standard","description":"Price-optimized models for rapid autocomplete, linting, high-frequency suggestions, and batch edits. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-04","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":120000,"output":65536},"status":"beta"},"text-standard":{"id":"text-standard","name":"Text Standard","description":"Price-optimized models with low-latency, high-throughput and shorter maximum outputs. Any model that meets the contract spec can serve your request.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-26","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":120000,"output":65536}},"gemini-pro-latest":{"id":"gemini-pro-latest","name":"Gemini Pro Latest","description":"Lab latest market for Gemini Pro frontier supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":65536}},"deepseek-pro-latest":{"id":"deepseek-pro-latest","name":"DeepSeek Pro Latest","description":"Lab latest market for DeepSeek Pro supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":384000}},"text-max":{"id":"text-max","name":"Text Max","description":"Frontier models for deep reasoning, long context, and complex workflows. Any model that meets the contract spec can serve your request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000}},"glm-latest":{"id":"glm-latest","name":"GLM Latest","description":"Lab latest market for GLM supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":917504,"output":131072}},"claude-opus-latest":{"id":"claude-opus-latest","name":"Claude Opus Latest","description":"Lab latest market for Claude Opus frontier supply. You contract for the latest qualifying route, not a specific model.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000}},"bytedance-pro-latest":{"id":"bytedance-pro-latest","name":"ByteDance Pro Latest","description":"Lab latest market for ByteDance Pro supply. You contract for the latest qualifying route, not a specific model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":256000,"output":131072}},"code-max":{"id":"code-max","name":"Code Max","description":"Frontier models for complex research, architectural decisions, debugging, and multi-file development. Any model that meets the contract spec can serve your request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-05-04","last_updated":"2026-07-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"status":"beta"}}},"tokengo":{"id":"tokengo","env":["TOKENGO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tokengo.com/v1","name":"TokenGo","doc":"https://www.tokengo.com/docs","models":{"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.89,"output":3.2647,"cache_read":0.2226},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.435,"output":0.87},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.19,"output":0.71,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":0.2174,"output":0.326,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.098,"output":0.196,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":2.65,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"}}},"neuralwatt":{"id":"neuralwatt","env":["NEURALWATT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.neuralwatt.com/v1","name":"Neuralwatt","doc":"https://portal.neuralwatt.com/docs","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.87,"output":1.74,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"gemma-4-31b":{"id":"gemma-4-31b","name":"Gemma 4 31B","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":16384},"cost":{"input":0.144,"output":0.42,"cache_read":0.0144},"canonical_model_id":"google/gemma-4-31b-it"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":262128},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.095},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"qwen-3.8-27b-flex":{"id":"qwen-3.8-27b-flex","name":"Qwen3.8 27B Flex","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":131072},"cost":{"input":0.2925,"output":2.08,"cache_read":0.1625},"canonical_model_id":"alibaba/qwen3.8-27b"},"kimi-k2.7-code-fast":{"id":"kimi-k2.7-code-fast","name":"Kimi K2.7 Code Fast","description":"Kimi K2.7 Code with reasoning capped to a short budget for lower latency; reasoning cannot be disabled on this model","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":262128},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.095},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"qwen3.6-35b-flex":{"id":"qwen3.6-35b-flex","name":"Qwen3.6 35B Flex","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131056,"output":131056},"cost":{"input":0.1885,"output":0.7475,"cache_read":0.01885},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"glm-5.3-flash-flex":{"id":"glm-5.3-flash-flex","name":"GLM-5.3 Flash Flex","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":0.0975,"output":0.325,"cache_read":0.0195},"canonical_model_id":"zhipuai/glm-5.3-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"qwen-3.8-27b":{"id":"qwen-3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":131072},"cost":{"input":0.45,"output":3.2,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-27b"},"kimi-k3-flex":{"id":"kimi-k3-flex","name":"Kimi K3 Flex","description":"Kimi K3 on the flex tier: discounted, best-effort latency, requests may be held under load","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":1.95,"output":9.75,"cache_read":0.195},"canonical_model_id":"moonshotai/kimi-k3"},"mimo-v2.6-pro-flex":{"id":"mimo-v2.6-pro-flex","name":"MiMo-V2.6-Pro Flex","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.5655,"output":1.131,"cache_read":0.0234},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"kimi-k3-fast":{"id":"kimi-k3-fast","name":"Kimi K3 Fast","description":"Kimi K3 with thinking disabled for low-latency tool calling, vision, and JSON work","family":"kimi-k3","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"glm-5.3":{"id":"glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":1.45,"output":4.5,"cache_read":0.145},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.6-35b-fast":{"id":"qwen3.6-35b-fast","name":"Qwen3.6 35B Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131056,"output":131056},"cost":{"input":0.29,"output":1.15,"cache_read":0.029},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"deepseek-v4.1-flash-flex":{"id":"deepseek-v4.1-flash-flex","name":"DeepSeek V4.1 Flash Flex","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.0975,"output":0.39,"cache_read":0.00975},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"kimi-k2.7-code-flex":{"id":"kimi-k2.7-code-flex","name":"Kimi K2.7 Code Flex","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262128,"output":262128},"status":"deprecated","cost":{"input":0.6175,"output":2.6,"cache_read":0.06175},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"qwen3.6-35b":{"id":"qwen3.6-35b","name":"Qwen3.6 35B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131056,"output":131056},"cost":{"input":0.29,"output":1.15,"cache_read":0.029},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"deepseek-v4-flash-speed":{"id":"deepseek-v4-flash-speed","name":"DeepSeek V4 Flash (Speed)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-v4-flash-flex":{"id":"deepseek-v4-flash-flex","name":"DeepSeek V4 Flash Flex","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.091,"output":0.182,"cache_read":0.0182},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-5.3-flex":{"id":"glm-5.3-flex","name":"GLM 5.3 Flex","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048560,"output":1048560},"cost":{"input":0.9425,"output":2.925,"cache_read":0.09425},"canonical_model_id":"zhipuai/glm-5.3"}}},"zenifra":{"id":"zenifra","env":["ZENIFRA_AI_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ai.zenifra.com/v1","name":"Zenifra","doc":"https://docs.zenifra.com","models":{"alibaba/qwen3.6-35b-a3b":{"id":"alibaba/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"shape":"completions"},"cost":{"input":0.19,"output":0.48},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"bailing":{"id":"bailing","env":["BAILING_API_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tbox.cn/api/llm/v1","name":"Bailing","doc":"https://alipaytbox.yuque.com/sxs0ba/ling/intro","models":{"Ring-1T":{"id":"Ring-1T","name":"Ring-1T","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"ring","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-10","last_updated":"2025-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.57,"output":2.29}},"Ling-1T":{"id":"Ling-1T","name":"Ling-1T","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-10","last_updated":"2025-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.57,"output":2.29}}}},"siliconflow-cn":{"id":"siliconflow-cn","env":["SILICONFLOW_CN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.siliconflow.cn/v1","name":"SiliconFlow (China)","doc":"https://cloud.siliconflow.com/models","models":{"Pro/moonshotai/Kimi-K2.6":{"id":"Pro/moonshotai/Kimi-K2.6","name":"Pro/moonshotai/Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.97140861,"output":4.0350819,"cache_read":0.16439223},"canonical_model_id":"moonshotai/kimi-k2.6"},"Pro/moonshotai/Kimi-K2.5":{"id":"Pro/moonshotai/Kimi-K2.5","name":"Pro/moonshotai/Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"status":"deprecated","cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"Pro/deepseek-ai/DeepSeek-R1":{"id":"Pro/deepseek-ai/DeepSeek-R1","name":"Pro/deepseek-ai/DeepSeek-R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":2.39115964,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-r1-0528"},"Pro/deepseek-ai/DeepSeek-V3":{"id":"Pro/deepseek-ai/DeepSeek-V3","name":"Pro/deepseek-ai/DeepSeek-V3","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.29889496,"output":1.19557982,"cache_read":0.0298895},"canonical_model_id":"deepseek/deepseek-v3-0324"},"Pro/deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"Pro/deepseek-ai/DeepSeek-V3.1-Terminus","name":"Pro/deepseek-ai/DeepSeek-V3.1-Terminus","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":1.79336973,"cache_read":0.05977899}},"Pro/deepseek-ai/DeepSeek-V3.2":{"id":"Pro/deepseek-ai/DeepSeek-V3.2","name":"Pro/deepseek-ai/DeepSeek-V3.2","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":0.89668487,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-v3.2"},"Pro/zai-org/GLM-5":{"id":"Pro/zai-org/GLM-5","name":"Pro/zai-org/GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":205000},"status":"deprecated","cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"Pro/zai-org/GLM-5.1":{"id":"Pro/zai-org/GLM-5.1","name":"Pro/zai-org/GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":205000},"cost":{"input":0.89668487,"output":3.58673947,"cache_read":0.19428172,"cache_write":0,"tiers":[{"input":1.19557982,"output":4.18452938,"cache_read":0.29889496,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"ByteDance-Seed/Seed-OSS-36B-Instruct":{"id":"ByteDance-Seed/Seed-OSS-36B-Instruct","name":"ByteDance-Seed/Seed-OSS-36B-Instruct","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-04","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.22417122,"output":0.59778991}},"tencent/Hunyuan-A13B-Instruct":{"id":"tencent/Hunyuan-A13B-Instruct","name":"tencent/Hunyuan-A13B-Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.59778991}},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen/Qwen3-8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen/Qwen3-VL-30B-A3B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen/Qwen3.5-27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.08966849,"output":0.71734789,"tiers":[{"input":0.26900546,"output":2.15204368,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-VL-32B-Thinking":{"id":"Qwen/Qwen3-VL-32B-Thinking","name":"Qwen/Qwen3-VL-32B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.14944748,"output":1.49447478}},"Qwen/Qwen3-VL-8B-Instruct":{"id":"Qwen/Qwen3-VL-8B-Instruct","name":"Qwen/Qwen3-VL-8B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.07472374,"output":0.29889496}},"Qwen/Qwen3-14B":{"id":"Qwen/Qwen3-14B","name":"Qwen/Qwen3-14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.07472374,"output":0.29889496}},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen/Qwen3-32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.59778991},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.5-4B":{"id":"Qwen/Qwen3.5-4B","name":"Qwen/Qwen3.5-4B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-4b"},"Qwen/Qwen3-VL-30B-A3B-Thinking":{"id":"Qwen/Qwen3-VL-30B-A3B-Thinking","name":"Qwen/Qwen3-VL-30B-A3B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-11","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294}},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen/Qwen3-Coder-30B-A3B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen/Qwen3-30B-A3B-Instruct-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.41845294},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen/Qwen3.5-35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.05977899,"output":0.47823193,"tiers":[{"input":0.23911596,"output":1.91292771,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen/Qwen3.6-35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26900546,"output":1.61403276},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-VL-32B-Instruct":{"id":"Qwen/Qwen3-VL-32B-Instruct","name":"Qwen/Qwen3-VL-32B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.14944748,"output":0.59778991}},"Qwen/Qwen2.5-7B-Instruct":{"id":"Qwen/Qwen2.5-7B-Instruct","name":"Qwen/Qwen2.5-7B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"Qwen/Qwen2.5-72B-Instruct":{"id":"Qwen/Qwen2.5-72B-Instruct","name":"Qwen/Qwen2.5-72B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0.61721808,"output":0.61721808}},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen/Qwen3.5-9B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.07472374,"output":0.59778991,"tiers":[{"input":0.22417122,"output":1.79336973,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen/Qwen3.5-122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.11955798,"output":0.95646386,"tiers":[{"input":0.29889496,"output":2.39115964,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"deepseek-ai/DeepSeek-R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":2.39115964,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"deepseek-ai/DeepSeek-V4-Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":393000},"cost":{"input":1.79336973,"output":3.58673947,"cache_read":0.14944748},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44834243,"output":1.3450273,"cache_read":0.04483424},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"deepseek-ai/DeepSeek-V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.29889496,"output":1.19557982,"cache_read":0.0298895},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"deepseek-ai/DeepSeek-V3.1-Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":1.79336973,"cache_read":0.05977899}},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"deepseek-ai/DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.59778991,"output":0.89668487,"cache_read":0.05977899},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-OCR":{"id":"deepseek-ai/DeepSeek-OCR","name":"deepseek-ai/DeepSeek-OCR","description":"OCR model for extracting structured text from documents and screenshots","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0}},"stepfun-ai/Step-3.5-Flash":{"id":"stepfun-ai/Step-3.5-Flash","name":"stepfun-ai/Step-3.5-Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.10461323,"output":0.3138397},"canonical_model_id":"stepfun/step-3.5-flash"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"zai-org/GLM-4.5-Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.89668487},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":1.19557982,"output":4.18452938,"cache_read":0.29889496,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"PaddlePaddle/PaddleOCR-VL-1.5":{"id":"PaddlePaddle/PaddleOCR-VL-1.5","name":"PaddlePaddle/PaddleOCR-VL-1.5","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-29","last_updated":"2026-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0,"output":0}},"inclusionAI/Ling-flash-2.0":{"id":"inclusionAI/Ling-flash-2.0","name":"inclusionAI/Ling-flash-2.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14944748,"output":0.59778991}}}},"runinfra":{"id":"runinfra","env":["RUNINFRA_GATEWAY_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.runinfra.ai/v1","name":"RunInfra","doc":"https://runinfra.ai/docs","models":{"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.15,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"ornith-ai/Ornith-1.5-35B-A3B":{"id":"ornith-ai/Ornith-1.5-35B-A3B","name":"Ornith 1.5 35B A3B","description":"Mixture-of-experts coding-reasoning model for agentic software tasks, tool use, and image understanding","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-18","last_updated":"2026-08-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"deepreinforce/ornith-1.5-35b-a3b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.13,"output":0.27,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.6,"output":1.9,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":1.3,"output":4.2,"cache_read":0.13},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.11,"output":0.45,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"Inferact/Qwen3.8-2.4T-A95B-NVFP4":{"id":"Inferact/Qwen3.8-2.4T-A95B-NVFP4","name":"Qwen3.8 2.4T A95B (NVFP4)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"}}},"xiaomi-token-plan-sgp":{"id":"xiaomi-token-plan-sgp","env":["XIAOMI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan-sgp.xiaomimimo.com/v1","name":"Xiaomi Token Plan (Singapore)","doc":"https://platform.xiaomimimo.com/#/docs","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"mimo-v2.5-tts":{"id":"mimo-v2.5-tts","type":"speech","name":"MiMo-V2.5-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mimo-v2.5-tts-voiceclone":{"id":"mimo-v2.5-tts-voiceclone","type":"speech","name":"MiMo-V2.5-TTS-VoiceClone","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voiceclone"},"mimo-v2.5-tts-voicedesign":{"id":"mimo-v2.5-tts-voicedesign","type":"speech","name":"MiMo-V2.5-TTS-VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.5-tts-voicedesign"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"}}},"perplexity":{"id":"perplexity","env":["PERPLEXITY_API_KEY"],"npm":"@ai-sdk/perplexity","name":"Perplexity","doc":"https://docs.perplexity.ai","models":{"sonar":{"id":"sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"sonar-reasoning-pro":{"id":"sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"sonar-deep-research":{"id":"sonar-deep-research","name":"Perplexity Sonar Deep Research","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"},"sonar-pro":{"id":"sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"}}},"ollama-cloud":{"id":"ollama-cloud","env":["OLLAMA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ollama.com/v1","name":"Ollama Cloud","doc":"https://docs.ollama.com/cloud","models":{"nemotron-3-ultra":{"id":"nemotron-3-ultra","name":"nemotron-3-ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.1,"output":3,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gpt-oss:120b":{"id":"gpt-oss:120b","name":"gpt-oss:120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.014},"canonical_model_id":"openai/gpt-oss-120b"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"kimi-k2.7-code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":976000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"deepseek-v4-pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"beam":{"id":"beam","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"minimax-m3":{"id":"minimax-m3","name":"minimax-m3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"},"deepseek-v4.1-flash:1017":{"id":"deepseek-v4.1-flash:1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"gpt-oss:20b":{"id":"gpt-oss:20b","name":"gpt-oss:20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.3,"cache_read":0.035},"canonical_model_id":"openai/gpt-oss-20b"},"nemotron-3-nano:30b":{"id":"nemotron-3-nano:30b","name":"nemotron-3-nano:30b","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"kimi-k3":{"id":"kimi-k3","name":"kimi-k3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"nemotron-3-super":{"id":"nemotron-3-super","name":"nemotron-3-super","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.015,"output":0.6,"cache_read":0.015},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"kimi-k2.6":{"id":"kimi-k2.6","name":"kimi-k2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax-m2.7":{"id":"minimax-m2.7","name":"minimax-m2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":196608},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"deepseek-v4-pro:0813":{"id":"deepseek-v4-pro:0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"gemma4:31b":{"id":"gemma4:31b","name":"gemma4:31b","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-31b-it"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"mistral-large-3:675b":{"id":"mistral-large-3:675b","name":"mistral-large-3:675b","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"muse-spark-open":{"id":"muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"}}},"scx-ai":{"id":"scx-ai","env":["SCX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.scx.ai/v1","name":"SCX.ai","doc":"https://platform.scx.ai/docs","models":{"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.17,"output":0.55},"canonical_model_id":"openai/gpt-oss-120b"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.55,"output":1.784,"cache_read":0.111},"canonical_model_id":"zhipuai/glm-5.2"},"Qwen3.8-Max":{"id":"Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":983616,"output":131072},"cost":{"input":1.815,"output":5.4461,"cache_read":0.17,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":196608},"cost":{"input":0.48,"output":1.79,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M2.7"}}},"ambient":{"id":"ambient","env":["AMBIENT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ambient.xyz/v1","name":"Ambient","doc":"https://ambient.xyz","models":{"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"cost":{"input":0.6,"output":2,"cache_read":0.15,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"unsloth/qwen3.8-27b-gguf":{"id":"unsloth/qwen3.8-27b-gguf","name":"qwen3.8-27b-4q","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"unsloth","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":101072,"output":101072},"cost":{"input":0.001,"output":0.001,"cache_read":0,"cache_write":0}},"ambient/small":{"id":"ambient/small","name":"Ambient Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":101072,"output":101072},"cost":{"input":0.001,"output":0.001,"cache_read":0,"cache_write":0}},"ambient/large":{"id":"ambient/large","name":"Ambient Large","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"cost":{"input":0.6,"output":2,"cache_read":0.15,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131702},"cost":{"input":0.23,"output":0.96,"cache_read":0.05,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.32,"output":3.2,"cache_read":0.16,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B-A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.09,"output":0.3,"cache_read":0.09,"cache_write":0},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":1,"cache_read":0.15,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.32,"output":3.2,"cache_read":0.16,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-27b"},"zai-org/GLM-5.2-FP8":{"id":"zai-org/GLM-5.2-FP8","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"status":"deprecated","cost":{"input":1.2,"output":4.2,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"amazon-bedrock":{"id":"amazon-bedrock","env":["AWS_ACCESS_KEY_ID","AWS_SECRET_ACCESS_KEY","AWS_REGION","AWS_BEARER_TOKEN_BEDROCK"],"npm":"@ai-sdk/amazon-bedrock","name":"Amazon Bedrock","doc":"https://docs.aws.amazon.com/bedrock/latest/userguide/models-supported.html","models":{"in.openai.gpt-5.6-luna":{"id":"in.openai.gpt-5.6-luna","name":"GPT-5.6 Luna (India)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"us.meta.llama4-maverick-17b-instruct-v1:0":{"id":"us.meta.llama4-maverick-17b-instruct-v1:0","name":"Llama 4 Maverick 17B Instruct (US)","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":8192},"status":"deprecated","cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"zai.glm-4.7-flash":{"id":"zai.glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"au.anthropic.claude-sonnet-4-6":{"id":"au.anthropic.claude-sonnet-4-6","name":"AU Anthropic Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"global.xai.grok-4.7":{"id":"global.xai.grok-4.7","name":"Grok 4.7 (Global)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"mistral.mistral-large-3-675b-instruct":{"id":"mistral.mistral-large-3-675b-instruct","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"openai.gpt-6-luna":{"id":"openai.gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275}},"canonical_model_id":"openai/gpt-6-luna"},"openai.gpt-5.4":{"id":"openai.gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-06-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.75,"output":16.5,"cache_read":0.275},"canonical_model_id":"openai/gpt-5.4"},"openai.gpt-oss-safeguard-20b":{"id":"openai.gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"anthropic.claude-fable-5-1":{"id":"anthropic.claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"mistral.voxtral-small-24b-2507":{"id":"mistral.voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"status":"deprecated","cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"deepseek.v3-v1:0":{"id":"deepseek.v3-v1:0","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":81920},"status":"deprecated","cost":{"input":0.58,"output":1.68},"canonical_model_id":"deepseek/deepseek-v3.1"},"openai.gpt-6-astra":{"id":"openai.gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":66,"output":330,"cache_read":6.6,"cache_write":82.5},"provider":{"body":{"service_tier":"ultrafast"}}}}},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75,"tiers":[{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5}},"canonical_model_id":"openai/gpt-6-astra"},"openai.gpt-6.1-sol":{"id":"openai.gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":13.2,"output":66,"cache_read":0.66,"cache_write":16.5},"provider":{"body":{"service_tier":"ultrafast"}}}}},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6.1-sol"},"in.anthropic.claude-sonnet-5":{"id":"in.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (India)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"global.xai.grok-4.6":{"id":"global.xai.grok-4.6","name":"Grok 4.6 (Global)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"au.anthropic.claude-sonnet-5":{"id":"au.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (AU)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"global.openai.gpt-6.1-sol":{"id":"global.openai.gpt-6.1-sol","name":"GPT-6.1 Sol (Global)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":12,"output":60,"cache_read":0.6,"cache_write":15},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"apac.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"apac.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (APAC)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"anthropic.claude-opus-4-1-20250805-v1:0":{"id":"anthropic.claude-opus-4-1-20250805-v1:0","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"global.openai.gpt-5.6-sol":{"id":"global.openai.gpt-5.6-sol","name":"GPT-5.6 Sol (Global)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"deepseek.r1-v1:0":{"id":"deepseek.r1-v1:0","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"qwen.qwen3-vl-235b-a22b":{"id":"qwen.qwen3-vl-235b-a22b","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262000},"cost":{"input":0.53,"output":2.66},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"openai.gpt-6-sol":{"id":"openai.gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6-sol"},"eu.anthropic.claude-sonnet-5-5":{"id":"eu.anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5 (EU)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"amazon.nova-micro-v1:0":{"id":"amazon.nova-micro-v1:0","name":"Nova Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875,"cache_write":0.035},"canonical_model_id":"amazon/nova-micro"},"au.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"au.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (AU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"qwen.qwen3-next-80b-a3b":{"id":"qwen.qwen3-next-80b-a3b","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-11","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262000},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"eu.amazon.nova-micro-v1:0":{"id":"eu.amazon.nova-micro-v1:0","name":"Nova Micro (EU)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.04,"output":0.16,"cache_read":0.01,"cache_write":0.04},"canonical_model_id":"amazon/nova-micro"},"global.anthropic.claude-opus-4-6-v1":{"id":"global.anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6 (Global)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"au.anthropic.claude-opus-5-5":{"id":"au.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (AU)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"openai.gpt-oss-120b":{"id":"openai.gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/v1","shape":"responses"},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"amazon.nova-lite-v1:0":{"id":"amazon.nova-lite-v1:0","name":"Nova Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015,"cache_write":0.06},"canonical_model_id":"amazon/nova-lite"},"au.anthropic.claude-opus-4-7":{"id":"au.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (AU)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"au.anthropic.claude-haiku-5-5":{"id":"au.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (AU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"global.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"global.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (Global)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"global.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"global.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (Global)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"us.anthropic.claude-fable-5":{"id":"us.anthropic.claude-fable-5","name":"Claude Fable 5 (US)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"zai.glm-5":{"id":"zai.glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"openai.gpt-5.6-terra":{"id":"openai.gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"google.gemma-3-12b-it":{"id":"google.gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"global.anthropic.claude-opus-5-5":{"id":"global.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (Global)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"global.moonshotai.kimi-k3":{"id":"global.moonshotai.kimi-k3","name":"Kimi K3 (Global)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"moonshotai/kimi-k3"},"us.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"us.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (US)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"us.mistral.pixtral-large-2502-v1:0":{"id":"us.mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02) (US)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"eu.anthropic.claude-sonnet-4-6":{"id":"eu.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (EU)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic.claude-sonnet-5":{"id":"anthropic.claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"us.openai.gpt-6-astra":{"id":"us.openai.gpt-6-astra","name":"GPT-6 Astra (US)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":66,"output":330,"cache_read":6.6,"cache_write":82.5},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75,"tiers":[{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":22,"output":82.5,"cache_read":2.2,"cache_write":27.5}},"canonical_model_id":"openai/gpt-6-astra"},"mistral.voxtral-mini-3b-2507":{"id":"mistral.voxtral-mini-3b-2507","name":"Voxtral Mini 3B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"status":"deprecated","cost":{"input":0.04,"output":0.04},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"global.anthropic.claude-opus-5":{"id":"global.anthropic.claude-opus-5","name":"Claude Opus 5 (Global)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"global.anthropic.claude-sonnet-5":{"id":"global.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (Global)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"au.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"au.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (AU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"us.anthropic.claude-opus-4-1-20250805-v1:0":{"id":"us.anthropic.claude-opus-4-1-20250805-v1:0","name":"Claude Opus 4.1 (US)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"jp.anthropic.claude-opus-4-7":{"id":"jp.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (JP)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"ca.amazon.nova-lite-v1:0":{"id":"ca.amazon.nova-lite-v1:0","name":"Nova Lite (CA)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.064,"output":0.256,"cache_read":0.016,"cache_write":0.064},"canonical_model_id":"amazon/nova-lite"},"global.openai.gpt-6-luna":{"id":"global.openai.gpt-6-luna","name":"GPT-6 Luna (Global)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"jp.anthropic.claude-sonnet-4-6":{"id":"jp.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (JP)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic.claude-opus-4-6-v1":{"id":"anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"us-gov.openai.gpt-oss-120b-1:0":{"id":"us-gov.openai.gpt-oss-120b-1:0","name":"gpt-oss-120b (GovCloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.18,"output":0.72},"canonical_model_id":"openai/gpt-oss-120b"},"minimax.minimax-m2.5":{"id":"minimax.minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":98304},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"},"global.openai.gpt-6-sol":{"id":"global.openai.gpt-6-sol","name":"GPT-6 Sol (Global)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"anthropic.claude-opus-4-5-20251101-v1:0":{"id":"anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"us.openai.gpt-6.1-sol":{"id":"us.openai.gpt-6.1-sol","name":"GPT-6.1 Sol (US)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":13.2,"output":66,"cache_read":0.66,"cache_write":16.5},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.22,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6.1-sol"},"eu.amazon.nova-lite-v1:0":{"id":"eu.amazon.nova-lite-v1:0","name":"Nova Lite (EU)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.069,"output":0.276,"cache_read":0.01725,"cache_write":0.069},"canonical_model_id":"amazon/nova-lite"},"moonshotai.kimi-k2.5":{"id":"moonshotai.kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-02-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"eu.anthropic.claude-opus-4-8":{"id":"eu.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (EU)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"openai.gpt-5.6-luna":{"id":"openai.gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"eu.anthropic.claude-sonnet-5":{"id":"eu.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (EU)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"us.meta.llama4-scout-17b-instruct-v1:0":{"id":"us.meta.llama4-scout-17b-instruct-v1:0","name":"Llama 4 Scout 17B Instruct (US)","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10000000,"output":8192},"status":"deprecated","cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"apac.amazon.nova-lite-v1:0":{"id":"apac.amazon.nova-lite-v1:0","name":"Nova Lite (APAC)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.063,"output":0.252,"cache_read":0.01575,"cache_write":0.063},"canonical_model_id":"amazon/nova-lite"},"google.gemma-3-27b-it":{"id":"google.gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.23,"output":0.38},"canonical_model_id":"google/gemma-3-27b-it"},"qwen.qwen3-235b-a22b-2507-v1:0":{"id":"qwen.qwen3-235b-a22b-2507-v1:0","name":"Qwen3 235B-A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.22,"output":0.88},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"global.anthropic.claude-opus-4-8":{"id":"global.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (Global)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic.claude-opus-5":{"id":"anthropic.claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"minimax.minimax-m2.1":{"id":"minimax.minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"xai.grok-4.6":{"id":"xai.grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":2.2,"output":6.6,"cache_read":0.55},"canonical_model_id":"xai/grok-4.6"},"anthropic.claude-haiku-5-5":{"id":"anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"moonshot.kimi-k2-thinking":{"id":"moonshot.kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16000},"status":"deprecated","cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"xai.grok-4.3":{"id":"xai.grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-06-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"us.moonshotai.kimi-k3":{"id":"us.moonshotai.kimi-k3","name":"Kimi K3 (US)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"moonshotai/kimi-k3"},"us.anthropic.claude-opus-4-6-v1":{"id":"us.anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6 (US)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"mistral.ministral-3-14b-instruct":{"id":"mistral.ministral-3-14b-instruct","name":"Ministral 14B 3.0","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"us.writer.palmyra-x4-v1:0":{"id":"us.writer.palmyra-x4-v1:0","name":"Palmyra X4 (US)","description":"Enterprise language model for workflow automation, coding, data analysis, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-10-09","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":122880,"output":8192},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"writer/palmyra-x4"},"nvidia.nemotron-nano-3-30b":{"id":"nvidia.nemotron-nano-3-30b","name":"NVIDIA Nemotron Nano 3 30B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"us.openai.gpt-6-sol":{"id":"us.openai.gpt-6-sol","name":"GPT-6 Sol (US)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":16.5,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-6-sol"},"global.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"global.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (Global)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"eu.anthropic.claude-opus-5":{"id":"eu.anthropic.claude-opus-5","name":"Claude Opus 5 (EU)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"jp.anthropic.claude-opus-4-8":{"id":"jp.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (JP)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"meta.llama3-1-8b-instruct-v1:0":{"id":"meta.llama3-1-8b-instruct-v1:0","name":"Llama 3.1 8B Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"qwen.qwen3-coder-next":{"id":"qwen.qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"us.anthropic.claude-haiku-5-5":{"id":"us.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (US)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"eu.anthropic.claude-opus-4-7":{"id":"eu.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (EU)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"eu.mistral.pixtral-large-2502-v1:0":{"id":"eu.mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02) (EU)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"eu.anthropic.claude-opus-4-6-v1":{"id":"eu.anthropic.claude-opus-4-6-v1","name":"Claude Opus 4.6 (EU)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"us-gov.openai.gpt-oss-20b-1:0":{"id":"us-gov.openai.gpt-oss-20b-1:0","name":"gpt-oss-20b (GovCloud)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.084,"output":0.36},"canonical_model_id":"openai/gpt-oss-20b"},"us.zai.glm-5.3":{"id":"us.zai.glm-5.3","name":"GLM-5.3 (US)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.848,"output":5.808,"cache_read":0.3432,"cache_write":2.31},"canonical_model_id":"zhipuai/glm-5.3"},"us.xai.grok-4.7":{"id":"us.xai.grok-4.7","name":"Grok 4.7 (US)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.2,"output":6.6,"cache_read":0.55},"canonical_model_id":"xai/grok-4.7"},"mistral.ministral-3-3b-instruct":{"id":"mistral.ministral-3-3b-instruct","name":"Ministral 3 3B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"mistral/ministral-3-3b-instruct-2512"},"deepseek.v3.2":{"id":"deepseek.v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2026-02-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":81920},"cost":{"input":0.62,"output":1.85},"canonical_model_id":"deepseek/deepseek-v3.2"},"us.anthropic.claude-opus-5-5":{"id":"us.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (US)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"us.anthropic.claude-opus-4-5-20251101-v1:0":{"id":"us.anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5 (US)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"us.amazon.nova-lite-v1:0":{"id":"us.amazon.nova-lite-v1:0","name":"Nova Lite (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015,"cache_write":0.06},"canonical_model_id":"amazon/nova-lite"},"eu.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"eu.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"nvidia.nemotron-nano-9b-v2":{"id":"nvidia.nemotron-nano-9b-v2","name":"NVIDIA Nemotron Nano 9B v2","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-08-18","last_updated":"2025-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.06,"output":0.23},"canonical_model_id":"nvidia/nemotron-nano-9b-v2"},"openai.gpt-5.6-sol":{"id":"openai.gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5,"tiers":[{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11}},"canonical_model_id":"openai/gpt-5.6-sol"},"eu.anthropic.claude-haiku-5-5":{"id":"eu.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (EU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"amazon.nova-pro-v1:0":{"id":"amazon.nova-pro-v1:0","name":"Nova Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2,"cache_write":0.8},"canonical_model_id":"amazon/nova-pro"},"global.anthropic.claude-haiku-5-5":{"id":"global.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (Global)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"jp.anthropic.claude-opus-5":{"id":"jp.anthropic.claude-opus-5","name":"Claude Opus 5 (JP)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"global.anthropic.claude-sonnet-5-5":{"id":"global.anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5 (Global)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"au.anthropic.claude-opus-5":{"id":"au.anthropic.claude-opus-5","name":"Claude Opus 5 (AU)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"meta.llama4-maverick-17b-instruct-v1:0":{"id":"meta.llama4-maverick-17b-instruct-v1:0","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":8192},"status":"deprecated","cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"eu.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"eu.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (EU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"us.anthropic.claude-sonnet-4-20250514-v1:0":{"id":"us.anthropic.claude-sonnet-4-20250514-v1:0","name":"Claude Sonnet 4 (US)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"us.xai.grok-4.6":{"id":"us.xai.grok-4.6","name":"Grok 4.6 (US)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2.2,"output":6.6,"cache_read":0.55},"canonical_model_id":"xai/grok-4.6"},"us.amazon.nova-pro-v1:0":{"id":"us.amazon.nova-pro-v1:0","name":"Nova Pro (US)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2,"cache_write":0.8},"canonical_model_id":"amazon/nova-pro"},"qwen.qwen3-32b-v1:0":{"id":"qwen.qwen3-32b-v1:0","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-32b"},"us.openai.gpt-5.6-sol":{"id":"us.openai.gpt-5.6-sol","name":"GPT-5.6 Sol (US)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5,"tiers":[{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11}},"canonical_model_id":"openai/gpt-5.6-sol"},"google.gemma-3-4b-it":{"id":"google.gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"status":"deprecated","cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"us.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"us.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (US)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"us.openai.gpt-5.6-luna":{"id":"us.openai.gpt-5.6-luna","name":"GPT-5.6 Luna (US)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"us.anthropic.claude-sonnet-5-5":{"id":"us.anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5 (US)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.11,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"us.anthropic.claude-fable-5-1":{"id":"us.anthropic.claude-fable-5-1","name":"Claude Fable 5.1 (US)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":0.275,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5-1"},"jp.anthropic.claude-opus-5-5":{"id":"jp.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (JP)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"amazon.nova-2-lite-v1:0":{"id":"amazon.nova-2-lite-v1:0","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.0825,"cache_write":0.33},"canonical_model_id":"amazon/nova-2-lite"},"meta.llama3-3-70b-instruct-v1:0":{"id":"meta.llama3-3-70b-instruct-v1:0","name":"Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"au.anthropic.claude-opus-4-6-v1":{"id":"au.anthropic.claude-opus-4-6-v1","name":"AU Anthropic Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"us.meta.llama3-1-8b-instruct-v1:0":{"id":"us.meta.llama3-1-8b-instruct-v1:0","name":"Llama 3.1 8B Instruct (US)","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"eu.amazon.nova-pro-v1:0":{"id":"eu.amazon.nova-pro-v1:0","name":"Nova Pro (EU)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.92,"output":3.68,"cache_read":0.23,"cache_write":0.92},"canonical_model_id":"amazon/nova-pro"},"google.gemma-4-e2b":{"id":"google.gemma-4-e2b","name":"Gemma 4 E2B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-4-E2B-it"},"in.anthropic.claude-opus-5":{"id":"in.anthropic.claude-opus-5","name":"Claude Opus 5 (India)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"meta.llama3-1-70b-instruct-v1:0":{"id":"meta.llama3-1-70b-instruct-v1:0","name":"Llama 3.1 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"mistral.devstral-2-123b":{"id":"mistral.devstral-2-123b","name":"Devstral 2 123B","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-2512"},"us.amazon.nova-micro-v1:0":{"id":"us.amazon.nova-micro-v1:0","name":"Nova Micro (US)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875,"cache_write":0.035},"canonical_model_id":"amazon/nova-micro"},"us.amazon.nova-premier-v1:0":{"id":"us.amazon.nova-premier-v1:0","name":"Nova Premier (US)","description":"Multimodal model for complex analysis, long-context understanding, tool use, and model distillation","family":"nova","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-30","last_updated":"2025-04-30","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":10000},"status":"deprecated","cost":{"input":2.5,"output":12.5,"cache_read":0.625,"cache_write":2.5},"canonical_model_id":"amazon/nova-premier"},"meta.llama4-scout-17b-instruct-v1:0":{"id":"meta.llama4-scout-17b-instruct-v1:0","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10000000,"output":8192},"status":"deprecated","cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"minimax.minimax-m2":{"id":"minimax.minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204608,"output":128000},"status":"deprecated","cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"global.anthropic.claude-fable-5":{"id":"global.anthropic.claude-fable-5","name":"Claude Fable 5 (Global)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"google.gemma-4-31b":{"id":"google.gemma-4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"apac.amazon.nova-micro-v1:0":{"id":"apac.amazon.nova-micro-v1:0","name":"Nova Micro (APAC)","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.037,"output":0.148,"cache_read":0.00925,"cache_write":0.037},"canonical_model_id":"amazon/nova-micro"},"anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"us.amazon.nova-2-lite-v1:0":{"id":"us.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (US)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.0825,"cache_write":0.33},"canonical_model_id":"amazon/nova-2-lite"},"anthropic.claude-opus-4-8":{"id":"anthropic.claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"mistral.ministral-3-8b-instruct":{"id":"mistral.ministral-3-8b-instruct","name":"Ministral 3 8B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.15,"output":0.15},"canonical_model_id":"mistral/ministral-3-8b-instruct-2512"},"in.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"in.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (India)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"google.gemma-4-26b-a4b":{"id":"google.gemma-4-26b-a4b","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"global.anthropic.claude-sonnet-4-6":{"id":"global.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (Global)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"writer.palmyra-x4-v1:0":{"id":"writer.palmyra-x4-v1:0","name":"Palmyra X4","description":"Enterprise language model for workflow automation, coding, data analysis, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-10-09","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":122880,"output":8192},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"writer/palmyra-x4"},"in.openai.gpt-5.6-terra":{"id":"in.openai.gpt-5.6-terra","name":"GPT-5.6 Terra (India)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai.gpt-5.5":{"id":"openai.gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-06-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/openai/v1","shape":"responses"},"cost":{"input":5.5,"output":33,"cache_read":0.55},"canonical_model_id":"openai/gpt-5.5"},"us.meta.llama3-3-70b-instruct-v1:0":{"id":"us.meta.llama3-3-70b-instruct-v1:0","name":"Llama 3.3 70B Instruct (US)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"global.zai.glm-5.3":{"id":"global.zai.glm-5.3","name":"GLM-5.3 (Global)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.68,"output":5.28,"cache_read":0.312,"cache_write":2.1},"canonical_model_id":"zhipuai/glm-5.3"},"eu.anthropic.claude-opus-5-5":{"id":"eu.anthropic.claude-opus-5-5","name":"Claude Opus 5.5 (EU)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic.claude-sonnet-4-6":{"id":"anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"global.anthropic.claude-opus-4-7":{"id":"global.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (Global)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"global.openai.gpt-5.6-terra":{"id":"global.openai.gpt-5.6-terra","name":"GPT-5.6 Terra (Global)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"jp.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"jp.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (JP)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"au.anthropic.claude-opus-4-8":{"id":"au.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (AU)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"openai.gpt-oss-safeguard-120b":{"id":"openai.gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"us.writer.palmyra-x5-v1:0":{"id":"us.writer.palmyra-x5-v1:0","name":"Palmyra X5 (US)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"input":1040000,"output":8192},"status":"deprecated","cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"eu.amazon.nova-2-lite-v1:0":{"id":"eu.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (EU)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.374,"output":3.157,"cache_read":0.0935,"cache_write":0.374},"canonical_model_id":"amazon/nova-2-lite"},"us.openai.gpt-5.6-terra":{"id":"us.openai.gpt-5.6-terra","name":"GPT-5.6 Terra (US)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"zai.glm-4.7":{"id":"zai.glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"mistral.pixtral-large-2502-v1:0":{"id":"mistral.pixtral-large-2502-v1:0","name":"Pixtral Large (25.02)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-2502"},"eu.anthropic.claude-opus-4-5-20251101-v1:0":{"id":"eu.anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5 (EU)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"jp.anthropic.claude-haiku-5-5":{"id":"jp.anthropic.claude-haiku-5-5","name":"Claude Haiku 5.5 (JP)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.055,"cache_write":0.6875,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"eu.anthropic.claude-sonnet-4-5-20250929-v1:0":{"id":"eu.anthropic.claude-sonnet-4-5-20250929-v1:0","name":"Claude Sonnet 4.5 (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"jp.amazon.nova-2-lite-v1:0":{"id":"jp.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (JP)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.396,"output":3.311,"cache_read":0.099,"cache_write":0.396},"canonical_model_id":"amazon/nova-2-lite"},"apac.amazon.nova-pro-v1:0":{"id":"apac.amazon.nova-pro-v1:0","name":"Nova Pro (APAC)","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.84,"output":3.36,"cache_read":0.21,"cache_write":0.84},"canonical_model_id":"amazon/nova-pro"},"us.anthropic.claude-opus-4-7":{"id":"us.anthropic.claude-opus-4-7","name":"Claude Opus 4.7 (US)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"qwen.qwen3-coder-30b-a3b-v1:0":{"id":"qwen.qwen3-coder-30b-a3b-v1:0","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"us.openai.gpt-6-luna":{"id":"us.openai.gpt-6-luna","name":"GPT-6 Luna (US)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.011,"cache_write":0.1375,"tiers":[{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.22,"output":0.825,"cache_read":0.022,"cache_write":0.275}},"canonical_model_id":"openai/gpt-6-luna"},"nvidia.nemotron-super-3-120b":{"id":"nvidia.nemotron-super-3-120b","name":"NVIDIA Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.15,"output":0.65},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"us.meta.llama3-1-70b-instruct-v1:0":{"id":"us.meta.llama3-1-70b-instruct-v1:0","name":"Llama 3.1 70B Instruct (US)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"anthropic.claude-opus-4-7":{"id":"anthropic.claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic.claude-opus-5-5":{"id":"anthropic.claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"global.anthropic.claude-opus-4-5-20251101-v1:0":{"id":"global.anthropic.claude-opus-4-5-20251101-v1:0","name":"Claude Opus 4.5 (Global)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"openai.gpt-oss-120b-1:0":{"id":"openai.gpt-oss-120b-1:0","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":128000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"us.anthropic.claude-sonnet-4-6":{"id":"us.anthropic.claude-sonnet-4-6","name":"Claude Sonnet 4.6 (US)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"openai.gpt-oss-20b-1:0":{"id":"openai.gpt-oss-20b-1:0","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":128000},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"global.anthropic.claude-fable-5-1":{"id":"global.anthropic.claude-fable-5-1","name":"Claude Fable 5.1 (Global)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"us.anthropic.claude-opus-5":{"id":"us.anthropic.claude-opus-5","name":"Claude Opus 5 (US)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"us.deepseek.r1-v1:0":{"id":"us.deepseek.r1-v1:0","name":"DeepSeek-R1 (US)","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"qwen.qwen3-coder-480b-a35b-v1:0":{"id":"qwen.qwen3-coder-480b-a35b-v1:0","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0.45,"output":1.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"us.anthropic.claude-sonnet-5":{"id":"us.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (US)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"global.amazon.nova-2-lite-v1:0":{"id":"global.amazon.nova-2-lite-v1:0","name":"Nova 2 Lite (Global)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.075,"cache_write":0.3},"canonical_model_id":"amazon/nova-2-lite"},"global.openai.gpt-6-astra":{"id":"global.openai.gpt-6-astra","name":"GPT-6 Astra (Global)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"ultrafast":{"cost":{"input":60,"output":300,"cache_read":6,"cache_write":75},"provider":{"body":{"additionalModelRequestFields":{"service_tier":"ultrafast"}}}}}},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"global.openai.gpt-5.6-luna":{"id":"global.openai.gpt-5.6-luna","name":"GPT-5.6 Luna (Global)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"eu.anthropic.claude-fable-5":{"id":"eu.anthropic.claude-fable-5","name":"Claude Fable 5 (EU)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic.claude-sonnet-5-5":{"id":"anthropic.claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"nvidia.nemotron-nano-12b-v2":{"id":"nvidia.nemotron-nano-12b-v2","name":"NVIDIA Nemotron Nano 12B v2 VL BF16","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.2,"output":0.6},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"anthropic.claude-fable-5":{"id":"anthropic.claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"openai.gpt-oss-20b":{"id":"openai.gpt-oss-20b","name":"gpt-oss-20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/amazon-bedrock/mantle","api":"https://bedrock-mantle.${AWS_REGION}.api.aws/v1","shape":"responses"},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"us.anthropic.claude-opus-4-8":{"id":"us.anthropic.claude-opus-4-8","name":"Claude Opus 4.8 (US)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"jp.anthropic.claude-haiku-4-5-20251001-v1:0":{"id":"jp.anthropic.claude-haiku-4-5-20251001-v1:0","name":"Claude Haiku 4.5 (JP)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"jp.anthropic.claude-sonnet-5":{"id":"jp.anthropic.claude-sonnet-5","name":"Claude Sonnet 5 (JP)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"mistral.magistral-small-2509":{"id":"mistral.magistral-small-2509","name":"Magistral Small 1.2","description":"Open multimodal reasoning model for transparent analysis of text and images","family":"magistral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-18","last_updated":"2025-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":40000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/magistral-small-2509"},"writer.palmyra-x5-v1:0":{"id":"writer.palmyra-x5-v1:0","name":"Palmyra X5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"palmyra","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"input":1040000,"output":8192},"status":"deprecated","cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"}}},"klokintegration":{"id":"klokintegration","env":["KLOKINTEGRATION_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api-gw.klok.ipaas.se/proxy/kloker-key/v1","name":"klokintegration.se","doc":"https://klokintegration.se/docs/ai-api","models":{"Kloker-Integration-Developer":{"id":"Kloker-Integration-Developer","name":"Kloker Integration Developer","description":"Knows the customer integration environment and Klok best practices. Opinionated about implementation. The gateway runs ecosystem lookup tools server-side and appends a system-prompt injection. Client system prompts and OpenAI tool calls are preserved.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":50000},"status":"beta","cost":{"input":0.23,"output":1.16}},"Kloker":{"id":"Kloker","name":"Kloker","description":"Cheap general model with a clean context. Nothing from the customer environment is packed in. It tracks the current best open source model. The Klok team verifies it and upgrades it periodically.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":50000},"status":"beta","cost":{"input":0.23,"output":1.16}},"Kloker-Integration-Architect":{"id":"Kloker-Integration-Architect","name":"Kloker Integration Architect","description":"Knows the customer integration environment and Klok best practices. Opinionated about structure. The gateway runs ecosystem lookup tools server-side and appends a system-prompt injection (data contracts, CloudEvents, event-driven flows). Client system prompts and OpenAI tool calls are preserved.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":50000},"status":"beta","cost":{"input":0.23,"output":1.16}}}},"sarvam":{"id":"sarvam","env":["SARVAM_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.sarvam.ai/v1","name":"Sarvam AI","doc":"https://docs.sarvam.ai/api-reference-docs/getting-started/models","models":{"sarvam-30b":{"id":"sarvam-30b","name":"Sarvam-30B","description":"Efficient Indian-language reasoning model for chat, coding, and multilingual work","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"status":"deprecated","canonical_model_id":"sarvam/sarvam-30b"},"sarvam-105b":{"id":"sarvam-105b","name":"Sarvam-105B","description":"Flagship Indian-language reasoning model for enterprise multilingual applications","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.15790269,"output":0.63161075,"cache_read":0.05263423},"canonical_model_id":"sarvam/sarvam-105b"}}},"alibaba":{"id":"alibaba","env":["DASHSCOPE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://dashscope-intl.aliyuncs.com/compatible-mode/v1","name":"Alibaba","doc":"https://www.alibabacloud.com/help/en/model-studio/models","models":{"qwen-turbo":{"id":"qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":16384},"status":"deprecated","cost":{"input":0.05,"output":0.2,"reasoning":0.5}},"qwen3-livetranslate-flash-realtime":{"id":"qwen3-livetranslate-flash-realtime","type":"realtime","name":"Qwen3-LiveTranslate Flash Realtime","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":53248,"output":4096},"status":"deprecated","provider":{"api":"wss://dashscope-intl.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":10,"output":10,"input_audio":10,"output_audio":38}},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.28,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5}}},"qvq-max":{"id":"qvq-max","name":"QVQ Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qvq","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":1.2,"output":4.8}},"qwen-plus":{"id":"qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.2,"reasoning":4,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":3.6,"reasoning":12,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":3.6,"reasoning":12,"cache_read":0.24,"cache_write":1.5}}},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.45,"output":2.25,"tiers":[{"input":0.75,"output":3.75,"tier":{"type":"context","size":32000}},{"input":1.2,"output":6,"tier":{"type":"context","size":128000}}]}},"qwen-mt-plus":{"id":"qwen-mt-plus","name":"Qwen-MT Plus","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":2.46,"output":7.37}},"qwen-max":{"id":"qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4}},"qwen-mt-turbo":{"id":"qwen-mt-turbo","name":"Qwen-MT Turbo","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"status":"deprecated","cost":{"input":0.16,"output":0.49}},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6}},"qwen-omni-turbo":{"id":"qwen-omni-turbo","name":"Qwen-Omni Turbo","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-19","last_updated":"2025-03-26","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.07,"output":0.27,"input_audio":4.44,"output_audio":8.89}},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":1.6,"cache_read":0.04,"cache_write":0.25,"tiers":[{"input":0.3,"output":2.4,"cache_read":0.06,"cache_write":0.375,"tier":{"type":"context","size":32000}},{"input":0.6,"output":4.8,"cache_read":0.12,"cache_write":0.75,"tier":{"type":"context","size":128000}}]}},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3.8-omni-flash":{"id":"qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}}},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":2}},"qwen-plus-character-ja":{"id":"qwen-plus-character-ja","name":"Qwen Plus Character (Japanese)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":512},"cost":{"input":0.5,"output":1.4}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"qwen-vl-ocr":{"id":"qwen-vl-ocr","name":"Qwen-VL OCR","description":"OCR model for extracting structured text from documents and screenshots","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-10-28","last_updated":"2025-04-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":34096,"output":4096},"cost":{"input":0.07,"output":0.16}},"qwen3-omni-flash-realtime":{"id":"qwen3-omni-flash-realtime","type":"realtime","name":"Qwen3-Omni Flash Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"status":"deprecated","provider":{"api":"wss://dashscope-intl.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.52,"output":1.99,"input_audio":4.57,"output_audio":18.13}},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.088},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.5-flash":{"id":"qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":81920}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"reasoning":2.4,"tiers":[{"input":0.5,"output":3,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.5,"output":3}}},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6}},"qwen3.5-122b-a10b":{"id":"qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2}},"qwen3-14b":{"id":"qwen3-14b","name":"Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.35,"output":1.4,"reasoning":4.2}},"qwen3-asr-flash":{"id":"qwen3-asr-flash","type":"transcription","name":"Qwen3-ASR Flash","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-04","release_date":"2025-09-08","last_updated":"2025-09-08","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":53248,"output":4096}},"qwen-vl-max":{"id":"qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.8,"output":3.2}},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.3125,"tiers":[{"input":1,"output":4,"cache_read":0.1,"cache_write":1.25,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1,"output":4,"cache_read":0.1,"cache_write":1.25}}},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.5,"output":7.5,"tiers":[{"input":2.7,"output":13.5,"tier":{"type":"context","size":32000}},{"input":4.5,"output":22.5,"tier":{"type":"context","size":128000}}]}},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.2,"output":6}},"qwen3-vl-30b-a3b":{"id":"qwen3-vl-30b-a3b","name":"Qwen3-VL 30B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.8,"reasoning":2.4}},"qwen-omni-turbo-realtime":{"id":"qwen-omni-turbo-realtime","type":"realtime","name":"Qwen-Omni Turbo Realtime","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-05-08","last_updated":"2025-05-08","modalities":{"input":["text","image","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048},"status":"deprecated","provider":{"api":"wss://dashscope-intl.aliyuncs.com/api-ws/v1/realtime"},"cost":{"input":0.27,"output":1.07,"input_audio":4.44,"output_audio":8.89}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1,"output":5}},"qwq-plus":{"id":"qwq-plus","name":"QwQ Plus","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.8,"output":2.4}},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":2.8,"reasoning":8.4}},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125}},"qwen3-next-80b-a3b-thinking":{"id":"qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":6}},"qwen-flash":{"id":"qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.05,"output":0.4,"cache_read":0.01,"cache_write":0.0625,"tiers":[{"input":0.25,"output":2,"cache_read":0.05,"cache_write":0.3125,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.25,"output":2,"cache_read":0.05,"cache_write":0.3125}}},"qwen2-5-omni-7b":{"id":"qwen2-5-omni-7b","name":"Qwen2.5-Omni 7B","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-12","last_updated":"2024-12","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":true,"limit":{"context":32768,"output":2048},"status":"deprecated","cost":{"input":0.1,"output":0.4,"input_audio":6.76}},"qwen-vl-plus":{"id":"qwen-vl-plus","name":"Qwen-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.21,"output":0.63}},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen3 235B-A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":2.8,"reasoning":8.4}},"qwen3-omni-flash":{"id":"qwen3-omni-flash","name":"Qwen3-Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":65536,"output":16384},"cost":{"input":0.43,"output":1.66,"input_audio":3.81,"output_audio":15.11}},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.375,"output":2.25}},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen3.6-max-preview":{"id":"qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.3,"output":7.8,"cache_read":0.13,"cache_write":1.625}},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2}},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4}},"qwen3-8b":{"id":"qwen3-8b","name":"Qwen3 8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.18,"output":0.7,"reasoning":2.1}},"qwen3-coder-flash":{"id":"qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.1,"cache_write":0.625,"tier":{"type":"context","size":32000}},{"input":0.8,"output":4,"cache_read":0.16,"cache_write":1,"tier":{"type":"context","size":128000}},{"input":1.6,"output":9.6,"cache_read":0.32,"cache_write":2,"tier":{"type":"context","size":256000}}]}},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":131072},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.0375,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"Qwen3-VL 235B-A22B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04","last_updated":"2025-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.7,"output":2.8,"reasoning":8.4}}}},"blueclaw":{"id":"blueclaw","env":["BLUECLAW_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://openai.blueclaw.network/v1","name":"Blue Claw","doc":"https://blueclaw.network","models":{"Qwen3.8-27B":{"id":"Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"beta","canonical_model_id":"alibaba/qwen3.8-27b"}}},"arcee":{"id":"arcee","env":["ARCEE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.arcee.ai/api/v1","name":"Arcee","doc":"https://docs.arcee.ai","models":{"trinity-large-thinking":{"id":"trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":80000},"status":"beta","cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"beta","cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"beta","cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":384000},"status":"deprecated","cost":{"input":1.74,"output":3.48,"cache_read":0.2},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"beta","cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-latest":{"id":"deepseek/deepseek-v4-flash-latest","name":"DeepSeek V4 Flash Latest","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"beta","cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"beta","cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"status":"beta","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-5.3":{"id":"zai-org/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"beta","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/glm-5.3-flash":{"id":"zai-org/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"beta","cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"upstage":{"id":"upstage","env":["UPSTAGE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.upstage.ai/v1/solar","name":"Upstage","doc":"https://developers.upstage.ai/docs/apis/chat","models":{"solar-pro2":{"id":"solar-pro2","name":"solar-pro2","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"cost":{"input":0.25,"output":0.25},"canonical_model_id":"upstage/solar-pro2"},"solar-mini":{"id":"solar-mini","name":"solar-mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"solar-mini","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-06-12","last_updated":"2025-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"cost":{"input":0.15,"output":0.15}},"solar-pro3":{"id":"solar-pro3","name":"solar-pro3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.25,"output":0.25},"canonical_model_id":"upstage/solar-pro3"},"solar-mini4":{"id":"solar-mini4","name":"Solar Mini 4","description":"Upstage's compact, cost-efficient model for agentic use","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"upstage/solar-mini4"},"solar-pro4":{"id":"solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"upstage/solar-pro4"}}},"zeldoc":{"id":"zeldoc","env":["ZELDOC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.zeldoc.ai/v1","name":"Zeldoc","doc":"https://docs.zeldoc.ai","models":{"zdev":{"id":"zdev","name":"ZDev","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0}}}},"zhipuai-coding-plan":{"id":"zhipuai-coding-plan","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://open.bigmodel.cn/api/coding/paas/v4","name":"Zhipu AI Coding Plan","doc":"https://docs.bigmodel.cn/cn/coding-plan/overview","models":{"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v"},"glm-5.3-highspeed":{"id":"glm-5.3-highspeed","name":"GLM-5.3 Highspeed","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"cloudflare-ai-gateway":{"id":"cloudflare-ai-gateway","env":["CLOUDFLARE_API_TOKEN","CLOUDFLARE_ACCOUNT_ID","CLOUDFLARE_GATEWAY_ID"],"npm":"ai-gateway-provider","name":"Cloudflare AI Gateway","doc":"https://developers.cloudflare.com/ai-gateway/","models":{"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember-1","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-pro":{"id":"google/gemini-3.1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"cache_write":1.5},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-flash":{"id":"google/gemini-3-flash","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite"},"unbiased/pareto":{"id":"unbiased/pareto","name":"Pareto","description":"Blended multimodal model that runs multiple language models in parallel and returns a synthesized answer","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25},"canonical_model_id":"unbiased/pareto"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.16},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.16},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.16},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1040000,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"alibaba/qwen3.7-plus":{"id":"alibaba/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.32,"output":1.28,"cache_read":0.064},"canonical_model_id":"alibaba/qwen3.7-plus"},"alibaba/qwen3.5-397b-a17b":{"id":"alibaba/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"alibaba/qwen3-max":{"id":"alibaba/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"alibaba/qwen3.7-max":{"id":"alibaba/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":3.75,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.7-max"},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":5,"output":30},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.25,"output":5,"cache_read":0.625},"canonical_model_id":"openai/gpt-4o"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/openai"},"cost":{"input":2,"output":10,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://gateway.ai.cloudflare.com/v1/${CLOUDFLARE_ACCOUNT_ID}/${CLOUDFLARE_GATEWAY_ID}/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"minimax/m3.1-flash":{"id":"minimax/m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015,"cache_write":0.15},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"reflection-ai/beam-1.0-501b-a23b":{"id":"reflection-ai/beam-1.0-501b-a23b","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"}}},"hpc-ai":{"id":"hpc-ai","env":["HPC_AI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.hpc-ai.com/inference/v1","name":"HPC-AI","doc":"https://www.hpc-ai.com/doc/docs/quickstart/","models":{"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1002000,"output":128000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-7"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"output":195000},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-5.1":{"id":"zai-org/glm-5.1","name":"GLM 5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202000,"output":202000},"cost":{"input":0.615,"output":2.46,"cache_read":0.133},"canonical_model_id":"zhipuai/glm-5.1"}}},"scnet-token-plan":{"id":"scnet-token-plan","env":["SCNET_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.scnet.cn/api/llm/v1","name":"SCNet Token Plan","doc":"https://www.scnet.cn/ac/openapi/doc/2.0/moduleapi/plans/token-plan.html","models":{"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"DeepSeek-V4.1-Flash":{"id":"DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3"},"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"Kimi-K2.7-Code":{"id":"Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"DeepSeek-V4-Flash-0731":{"id":"DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"Kimi-K3":{"id":"Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k3"},"GLM-5.3":{"id":"GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3"},"GLM-5":{"id":"GLM-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.1"},"GLM-5.3-Flash":{"id":"GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.2"},"Qwen3.8-Max":{"id":"Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-max"},"Qwen3.8-Flash":{"id":"Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"DeepSeek-V4-Pro-0813":{"id":"DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"Kimi-K2.5":{"id":"Kimi-K2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"togetherai":{"id":"togetherai","env":["TOGETHER_API_KEY"],"npm":"@ai-sdk/togetherai","name":"Together AI","doc":"https://docs.together.ai/docs/serverless-models","models":{"nvidia/NVIDIA-Nemotron-3.5-Super-VL":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512300,"output":512300},"status":"deprecated","cost":{"input":0.6,"output":3.6,"cache_read":0.2},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":524288},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.5"},"LiquidAI/LFM2-24B-A2B":{"id":"LiquidAI/LFM2-24B-A2B","name":"LFM2-24B-A2B","description":"Efficient open Liquid MoE instruct model for on-device chat and tool calling","family":"liquid","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.03,"output":0.12},"canonical_model_id":"liquid/lfm2-24b-a2b"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":131072},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"reflection/Beam-1.0-501B-A23B":{"id":"reflection/Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":1.2,"cache_read":0.04},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"meta-llama/Llama-3.3-70B-Instruct-Turbo":{"id":"meta-llama/Llama-3.3-70B-Instruct-Turbo","name":"Llama 3.3 70B","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":1.04,"output":1.04},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"Qwen/Qwen2.5-7B-Instruct-Turbo":{"id":"Qwen/Qwen2.5-7B-Instruct-Turbo","name":"Qwen 2.5 7B Instruct Turbo","description":"Small open Qwen2.5 instruction model for chat, structured output, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.3,"output":0.3},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"Qwen/Qwen3.6-Plus":{"id":"Qwen/Qwen3.6-Plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"alibaba/qwen3.6-plus"},"Qwen/Qwen3-Coder-Next-FP8":{"id":"Qwen/Qwen3-Coder-Next-FP8","name":"Qwen3 Coder Next FP8","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":130000},"status":"deprecated","cost":{"input":0.6,"output":3.6,"cache_read":0.35},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":2,"output":2},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3-235B-A22B-Instruct-2507-tput":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507-tput","name":"Qwen3 235B A22B Instruct 2507 FP8","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.2,"output":0.6},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3.7-Max":{"id":"Qwen/Qwen3.7-Max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.7-max"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163839,"output":163839},"status":"deprecated","cost":{"input":3,"output":7},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":1.25,"output":1.25},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek-ai/DeepSeek-V3-1":{"id":"deepseek-ai/DeepSeek-V3-1","name":"DeepSeek V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":1.7},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":262144},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":400000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":164000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"meta-models/Muse-Spark-Open-816B-A42B":{"id":"meta-models/Muse-Spark-Open-816B-A42B","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"essentialai/Rnj-1-Instruct":{"id":"essentialai/Rnj-1-Instruct","name":"Rnj-1 Instruct","description":"Open 8B Essential AI model for code, STEM reasoning, and agentic tool use","family":"rnj","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-04","last_updated":"2025-12-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.15},"canonical_model_id":"essentialai/rnj-1-instruct"}}},"privatemode-ai":{"id":"privatemode-ai","env":["PRIVATEMODE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://localhost:8080/v1","name":"Privatemode AI","doc":"https://docs.privatemode.ai/api/overview","models":{"kimi-latest":{"id":"kimi-latest","name":"Kimi (latest)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"status":"deprecated","cost":{"input":1.791,"output":8.9436,"cache_read":0.1733},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.2311,"output":0.7511,"cache_read":0.0462},"canonical_model_id":"openai/gpt-oss-120b"},"glm-flash-latest":{"id":"glm-flash-latest","name":"GLM Flash (latest)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2311,"output":0.7511,"cache_read":0.0578},"canonical_model_id":"zhipuai/glm-5.3-flash"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"status":"deprecated","cost":{"input":1.791,"output":8.9436,"cache_read":0.1733},"canonical_model_id":"moonshotai/kimi-k2.6"},"voxtral-mini-3b":{"id":"voxtral-mini-3b","type":"transcription","name":"Voxtral Mini 3B","description":"Speech-to-text model for audio transcription, translation, and audio understanding","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.00462,"output":0},"canonical_model_id":"mistral/voxtral-mini-3b-2507"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.791,"output":6.6326,"cache_read":0.1733},"canonical_model_id":"zhipuai/glm-5.3"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2311,"output":0.7511,"cache_read":0.0578},"canonical_model_id":"zhipuai/glm-5.3-flash"},"deepseek-ocr-2":{"id":"deepseek-ocr-2","name":"DeepSeek OCR 2","description":"High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"status":"beta","cost":{"input":0.8897,"output":1.4675,"cache_read":0.0924},"canonical_model_id":"deepseek/deepseek-ocr-2"},"glm-latest":{"id":"glm-latest","name":"GLM (latest)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.791,"output":6.6326,"cache_read":0.1733},"canonical_model_id":"zhipuai/glm-5.3"},"whisper-large-v3":{"id":"whisper-large-v3","type":"transcription","name":"Whisper large-v3","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.01618,"output":0},"canonical_model_id":"openai/whisper-large-v3"},"qwen3-embedding-4b":{"id":"qwen3-embedding-4b","type":"embedding","name":"Qwen3-Embedding 4B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-06","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":2560},"cost":{"input":0.1502,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-4b"}}},"frogbot":{"id":"frogbot","env":["FROGBOT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://app.frogbot.ai/api/v1","name":"FrogBot","doc":"https://docs.frogbot.ai","models":{"minimax-m2-5":{"id":"minimax-m2-5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":8192},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"gpt-5-5":{"id":"gpt-5-5","name":"GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"gpt-5-4-mini":{"id":"gpt-5-4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"gemini-3-1-pro-preview":{"id":"gemini-3-1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek v4 Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-20b"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":8192},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"grok-4-3":{"id":"grok-4-3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.075},"canonical_model_id":"google/gemini-2.5-flash"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gpt-5-3-codex":{"id":"gpt-5-3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.2,"output":1.5,"cache_read":0.02},"canonical_model_id":"xai/grok-code-fast-1"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.31},"canonical_model_id":"google/gemini-2.5-pro"},"zai-glm-5-1":{"id":"zai-glm-5-1","name":"Z.AI GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":198000,"output":8192},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"gpt-5-4-nano":{"id":"gpt-5-4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"qwen-3-6-plus":{"id":"qwen-3-6-plus","name":"Qwen 3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.6-plus"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"}}},"liquid":{"id":"liquid","env":["LIQUID_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://router.inference.ai.exchange/v1","name":"Liquid Inference","doc":"https://inference.ai.exchange/docs","models":{"liquid.auto":{"id":"liquid.auto","name":"Liquid Auto","description":"Routes each request to the lowest-priced eligible offer on the Liquid Inference exchange","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"output":1048576}}}},"vivgrid":{"id":"vivgrid","env":["VIVGRID_API_KEY"],"npm":"@ai-sdk/openai","api":"https://api.vivgrid.com/v1","name":"Vivgrid","doc":"https://docs.vivgrid.com/models","models":{"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.2,"output":4.2,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.2"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.75,"output":3.75,"cache_read":0.15},"canonical_model_id":"google/gemini-3.8-flash"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.31,"output":1.23,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":3,"output":12,"cache_read":0.3,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.74,"output":3.48,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5-mini"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":1.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":10,"output":50,"cache_read":0.5,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.35,"output":3,"reasoning":3,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT 5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":1.2,"output":4.2,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.3"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT 5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":1,"output":6,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.28,"output":0.42,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.44,"output":1.32,"reasoning":1.32,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/openai-compatible"},"cost":{"input":0.15,"output":0.5,"cache_read":0.04},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT 5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"openai/gpt-5.6-sol"},"viv-fast":{"id":"viv-fast","name":"Viv Fast","description":"Fast coding model","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-09","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":256000},"cost":{"input":0.13,"output":0.4,"cache_read":0.05},"canonical_model_id":"vivgrid/viv-fast"}}},"jalapeno":{"id":"jalapeno","env":["JALAPENO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.jalapeno-cloud.ai/v1","name":"Jalapeno Cloud","doc":"https://www.jalapeno-cloud.ai/docs/","models":{"Qwen3.5-27B":{"id":"Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen3-VL-235B-A22B-Instruct":{"id":"Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":129024,"output":32768},"status":"deprecated","cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"Qwen3-Next-80B-A3B-Thinking":{"id":"Qwen3-Next-80B-A3B-Thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.6,"output":3.38,"cache_read":0.14},"canonical_model_id":"deepseek/deepseek-v4-pro"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"Kimi-K2.7-Code":{"id":"Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":271360,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":129024,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"Qwen3-VL-235B-A22B-Thinking":{"id":"Qwen3-VL-235B-A22B-Thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"Kimi-K3":{"id":"Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"Qwen3.5-35B-A3B":{"id":"Qwen3.5-35B-A3B","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"Hy3":{"id":"Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"input":192000,"output":128000},"status":"deprecated","cost":{"input":0.14,"output":0.58},"canonical_model_id":"tencent/hy3"},"Qwen3.5-122B-A10B":{"id":"Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"Qwen3.5-397B-A17B":{"id":"Qwen3.5-397B-A17B","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Kimi-K2.5":{"id":"Kimi-K2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":180224},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"}}},"kuae-cloud-coding-plan":{"id":"kuae-cloud-coding-plan","env":["KUAE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://coding-plan-endpoint.kuaecloud.net/v1","name":"KUAE Cloud Coding Plan","doc":"https://docs.mthreads.com/kuaecloud/kuaecloud-doc-online/coding_plan/","models":{"GLM-4.7":{"id":"GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"}}},"modelscope":{"id":"modelscope","env":["MODELSCOPE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api-inference.modelscope.cn/v1","name":"ModelScope","doc":"https://modelscope.cn/docs/model-service/API-Inference/intro","models":{"ZhipuAI/GLM-4.5":{"id":"ZhipuAI/GLM-4.5","name":"GLM-4.5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.5"},"ZhipuAI/GLM-4.6":{"id":"ZhipuAI/GLM-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.6"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3-235B-A22B-Thinking-2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3-30B-A3B-Thinking-2507":{"id":"Qwen/Qwen3-30B-A3B-Thinking-2507","name":"Qwen3 30B A3B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}}}},"hyper":{"id":"hyper","env":["HYPER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://hyper.charm.land/v1","name":"Charm Hyper","doc":"https://hyper.charm.land","models":{"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16000},"cost":{"input":1.03436,"output":4.3552,"cache_read":0.206872},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":13107},"cost":{"input":0.178,"output":0.68,"cache_read":0.089},"canonical_model_id":"openai/gpt-oss-120b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"cost":{"input":1.52432,"output":4.79072,"cache_read":0.152432},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.2,"output":4.8,"cache_read":0.24},"canonical_model_id":"alibaba/qwen3.7-plus"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.33,"output":1.31,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0.33,"output":1.31,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"cost":{"input":0.32664,"output":1.30656,"cache_read":0.064239},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"prism":{"id":"prism","name":"Prism","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":true,"release_date":"1970-01-01","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":16000},"cost":{"input":3.2664,"output":16.332,"cache_read":0.32664}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16000},"cost":{"input":3.2664,"output":16.332,"cache_read":0.32664},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":1.437216,"output":4.311648,"cache_read":0.047907},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":6553},"cost":{"input":0.484,"output":1.852,"cache_read":0.242},"canonical_model_id":"minimax/MiniMax-M2.7"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.52432,"output":4.79072,"cache_read":0.283088},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.5,"output":7.5,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.7-max"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":26214},"cost":{"input":0.6,"output":2.5,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":25600},"cost":{"input":0.098,"output":0.334,"cache_read":0.049},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.16332,"output":0.5444,"cache_read":0.031575},"canonical_model_id":"zhipuai/glm-5.3-flash"},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":1.0888,"output":4.40964,"cache_read":0.185096},"canonical_model_id":"thinkingmachines/inkling"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":64000},"cost":{"input":0.2,"output":0.8,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.7-flash"}}},"lowrouter":{"id":"lowrouter","env":["LOWROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lowrouter.ai/v1","name":"LowRouter","doc":"https://lowrouter.ai/docs/models/available","models":{"auto/mistralai/ministral-14b-2512":{"id":"auto/mistralai/ministral-14b-2512","name":"Ministral 3 14B","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2092,"output":0.2092,"cache_read":0.0209},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"auto/mistralai/codestral-2508":{"id":"auto/mistralai/codestral-2508","name":"Codestral 25.08","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.3138,"output":0.9413,"cache_read":0.0314},"canonical_model_id":"mistral/codestral-2508"},"auto/mistralai/ministral-8b-2512":{"id":"auto/mistralai/ministral-8b-2512","name":"Ministral 3 8B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1569,"output":0.1569,"cache_read":0.0157},"canonical_model_id":"mistral/ministral-3-8b-instruct-2512"},"auto/mistralai/mistral-large-4":{"id":"auto/mistralai/mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"status":"beta","cost":{"input":1.4367,"output":4.4156,"cache_read":0.1479},"canonical_model_id":"mistral/mistral-large-4"},"auto/mistralai/mistral-small-2603":{"id":"auto/mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.1569,"output":0.6275,"cache_read":0.0157},"canonical_model_id":"mistral/mistral-small-2603"},"auto/mistralai/mistral-large-2512":{"id":"auto/mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5229,"output":1.5688,"cache_read":0.0523},"canonical_model_id":"mistral/mistral-large-2512"},"auto/mistralai/mistral-medium-3.5":{"id":"auto/mistralai/mistral-medium-3.5","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5688,"output":7.8442,"cache_read":0.1569},"canonical_model_id":"mistral/mistral-medium-2604"},"auto/mistralai/ministral-3b-2512":{"id":"auto/mistralai/ministral-3b-2512","name":"Ministral 3 3B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1046,"output":0.1046,"cache_read":0.0105},"canonical_model_id":"mistral/ministral-3-3b-instruct-2512"},"auto/openai/gpt-4.1-mini":{"id":"auto/openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"auto/openai/gpt-oss-120b":{"id":"auto/openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1695,"output":0.6779},"canonical_model_id":"openai/gpt-oss-120b"},"auto/openai/gpt-oss-safeguard-120b":{"id":"auto/openai/gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.18,"output":0.73},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"auto/openai/gpt-5.4":{"id":"auto/openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"auto/openai/gpt-6-astra":{"id":"auto/openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"auto/openai/gpt-4.1":{"id":"auto/openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"auto/openai/gpt-oss-20b":{"id":"auto/openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.113,"output":0.4745},"canonical_model_id":"openai/gpt-oss-20b"},"auto/openai/gpt-5-mini":{"id":"auto/openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"auto/openai/gpt-5.5":{"id":"auto/openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"auto/openai/gpt-6-luna":{"id":"auto/openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"auto/openai/gpt-5-nano":{"id":"auto/openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"auto/openai/gpt-6-sol":{"id":"auto/openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"auto/openai/gpt-5.4-mini":{"id":"auto/openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"auto/openai/gpt-5.2":{"id":"auto/openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"auto/openai/gpt-5.6-terra":{"id":"auto/openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"auto/openai/gpt-6.1-sol":{"id":"auto/openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"auto/openai/gpt-oss-safeguard-20b":{"id":"auto/openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.24},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"auto/openai/gpt-4o":{"id":"auto/openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"auto/openai/o3":{"id":"auto/openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"auto/openai/gpt-5.6-luna":{"id":"auto/openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"auto/openai/gpt-4o-mini":{"id":"auto/openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"auto/openai/gpt-5":{"id":"auto/openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"auto/openai/gpt-5.6-sol":{"id":"auto/openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"}}},"trustedrouter":{"id":"trustedrouter","env":["TRUSTEDROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.trustedrouter.com/v1","name":"TrustedRouter","doc":"https://trustedrouter.com/docs","models":{"trustedrouter/synth-code":{"id":"trustedrouter/synth-code","name":"Synth Code","description":"TrustedRouter code synthesis orchestration alias that combines multiple model responses into one answer.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-20","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/e2e":{"id":"trustedrouter/e2e","name":"End-to-End Encrypted","description":"TrustedRouter privacy routing alias for end-to-end encrypted provider routes where available.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/fast":{"id":"trustedrouter/fast","name":"Fast","description":"TrustedRouter speed routing alias that prefers low-latency healthy model endpoints.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/synth":{"id":"trustedrouter/synth","name":"Synth","description":"TrustedRouter synthesis orchestration alias that combines multiple model responses into one answer.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-20","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/cheap":{"id":"trustedrouter/cheap","name":"Cheap","description":"TrustedRouter low-cost routing alias that prefers inexpensive healthy model endpoints.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/auto":{"id":"trustedrouter/auto","name":"Auto","description":"TrustedRouter automatic routing alias that chooses a healthy supported model endpoint for the request.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"trustedrouter/zdr":{"id":"trustedrouter/zdr","name":"Zero Data Retention","description":"TrustedRouter privacy routing alias that prefers zero data retention model endpoints.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}}}},"opencode":{"id":"opencode","env":["OPENCODE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://opencode.ai/inference/openai/v1","name":"OpenCode","doc":"https://opencode.ai/v2/docs/console/models/","models":{"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"ling-2.6-flash-free":{"id":"ling-2.6-flash-free","name":"Ling 2.6 Flash Free","description":"Legacy model retained for compatibility with older integrations","family":"ling-flash-free","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":32800},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.6,"output":2.2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"ll-alpha-free":{"id":"ll-alpha-free","name":"LL Alpha Free","description":"Best in class local model","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"claude-sonnet-4":{"id":"claude-sonnet-4","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"minimax-m2.5-free":{"id":"minimax-m2.5-free","name":"MiniMax-M2.5 Free","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.5"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"trinity-large-preview-free":{"id":"trinity-large-preview-free","name":"Trinity Large Preview","description":"Legacy model retained for compatibility with older integrations","family":"trinity","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-01-27","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0},"canonical_model_id":"arcee-ai/trinity-large-preview"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5.1"},"glm-4.7-free":{"id":"glm-4.7-free","name":"GLM-4.7 Free","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-4.7"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"qwen3-coder":{"id":"qwen3-coder","name":"Qwen3 Coder","description":"Legacy model retained for compatibility with older integrations","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.45,"output":1.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.8-flash"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"ling-3.0-tiny-free":{"id":"ling-3.0-tiny-free","name":"Ling-3.0-tiny Free","description":"Compact MoE model for responsive agents, instruction following, and multi-turn conversations","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.6,"output":2.2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.7"},"gpt-5.3-codex-spark":{"id":"gpt-5.3-codex-spark","name":"GPT-5.3 Codex Spark","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-spark","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex-spark"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5.1-codex"},"qwen3.6-plus-free":{"id":"qwen3.6-plus-free","name":"Qwen3.6 Plus Free","description":"Legacy model retained for compatibility with older integrations","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.6-plus"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.7-flash"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"north-mini-code-free":{"id":"north-mini-code-free","name":"North Mini Code Free","description":"Cohere coding model for practical software engineering and agentic edits","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"longcat-2.0-free":{"id":"longcat-2.0-free","name":"LongCat-2.0 Free","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meituan/longcat-2.0"},"deepseek-v4.1-flash-free":{"id":"deepseek-v4.1-flash-free","name":"DeepSeek V4.1 Flash Free","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"mimo-v2-omni-free":{"id":"mimo-v2-omni-free","name":"MiMo V2 Omni Free","description":"Legacy model retained for compatibility with older integrations","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2-omni"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"laguna-s-2.1-free":{"id":"laguna-s-2.1-free","name":"Laguna S 2.1 Free","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"poolside/laguna-s-2.1"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"step-5-preview-free":{"id":"step-5-preview-free","name":"Step 5 Preview Free","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"stepfun/step-5-preview"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":128000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"minimax-m3.1-flash-free":{"id":"minimax-m3.1-flash-free","name":"MiniMax M3.1 Flash Free","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"minimax-m2.1":{"id":"minimax-m2.1","name":"MiniMax-M2.1","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.1},"canonical_model_id":"minimax/MiniMax-M2.1"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"beam-free":{"id":"beam-free","name":"Beam Free","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"minimax-m2.1-free":{"id":"minimax-m2.1-free","name":"MiniMax-M2.1 Free","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M2.1"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"flint-40b-free":{"id":"flint-40b-free","name":"Flint 40B Free","description":"Open-weight mixture-of-experts model with about 2.5B active parameters for local coding agents","family":"flint","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06","release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"recursive/flint"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"kimi-k2.5-free":{"id":"kimi-k2.5-free","name":"Kimi K2.5 Free","description":"Legacy model retained for compatibility with older integrations","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"ling-3.0-flash-fin-free":{"id":"ling-3.0-flash-fin-free","name":"Ling 3.0 Flash Fin Free","description":"Finance-enhanced model for financial research, multi-step investment workflows, and long-horizon planning and execution","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-09-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.0-flash-fin"},"deepseek-v4-flash-free":{"id":"deepseek-v4-flash-free","name":"DeepSeek V4 Flash Free","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"nemotron-3-super-free":{"id":"nemotron-3-super-free","name":"Nemotron 3 Super Free","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":30,"output":180,"cache_read":30},"canonical_model_id":"openai/gpt-5.5-pro"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"glm-5-free":{"id":"glm-5-free","name":"GLM-5 Free","description":"Legacy model retained for compatibility with older integrations","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5"},"mimo-v2-pro-free":{"id":"mimo-v2-pro-free","name":"MiMo V2 Pro Free","description":"Legacy model retained for compatibility with older integrations","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2-pro"},"cat-jam-free":{"id":"cat-jam-free","name":"Cat Jam Free","description":"A text-only reasoning model for coding and agentic workflows, with tool calling and structured JSON outputs.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax-m3-free":{"id":"minimax-m3-free","name":"MiniMax-M3 Free","description":"Legacy model retained for compatibility with older integrations","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"minimax/MiniMax-M3"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"space-bunny-free":{"id":"space-bunny-free","name":"Space Bunny Free","description":"Anonymous preview reasoning model for coding, agentic tasks, tool use, and multimodal input","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":524288,"output":524288},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"claude-3-5-haiku":{"id":"claude-3-5-haiku","name":"Claude Haiku 3.5","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.8,"output":4,"cache_read":0.08,"cache_write":1},"canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"hy3-preview-free":{"id":"hy3-preview-free","name":"Hy3 preview Free","description":"Legacy model retained for compatibility with older integrations","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"tencent/hy3-preview"},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"alibaba/qwen3.5-plus"},"gemini-3-pro":{"id":"gemini-3-pro","name":"Gemini 3 Pro","description":"Legacy model retained for compatibility with older integrations","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3-pro-preview"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":30,"output":180,"cache_read":30},"canonical_model_id":"openai/gpt-5.4-pro"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"nemotron-3.5-lightning-free":{"id":"nemotron-3.5-lightning-free","name":"Nemotron 3.5 Lightning Free","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"provider":{"npm":"@ai-sdk/mistral","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"x-preview-f-free":{"id":"x-preview-f-free","name":"Ox Alpha Free (Unlimited)","description":"Stealth reasoning model for coding, agentic tasks, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-08-21","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0}},"ling-3.0-flash-free":{"id":"ling-3.0-flash-free","name":"Ling-3.0-flash Free","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.0-flash"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"mimo-v2-flash-free":{"id":"mimo-v2-flash-free","name":"MiMo V2 Flash Free","description":"Legacy model retained for compatibility with older integrations","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2-flash"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"ring-2.6-1t-free":{"id":"ring-2.6-1t-free","name":"Ring 2.6 1T Free","description":"Legacy model retained for compatibility with older integrations","family":"ring-1t-free","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-05-08","last_updated":"2026-05-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":66000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"nemotron-3-ultra-free":{"id":"nemotron-3-ultra-free","name":"Nemotron 3 Ultra Free","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"mimo-v2.6-flash-free":{"id":"mimo-v2.6-flash-free","name":"MiMo-V2.6-Flash Free","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"gemini-3.1-pro":{"id":"gemini-3.1-pro","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":1.5,"output":7.5,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.6-flash"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.5"},"mimo-v2.5-free":{"id":"mimo-v2.5-free","name":"MiMo V2.5 Free","description":"MiMo omni model for text, image, video, audio, and agents","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"fledge-alpha-free":{"id":"fledge-alpha-free","name":"Fledge Alpha Free","description":"Free-tier preview reasoning model with text and image input and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"nemotron-3.5-super-vl-free":{"id":"nemotron-3.5-super-vl-free","name":"Nemotron 3.5 Super VL Free","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"muse-spark-1.3-contributor-free":{"id":"muse-spark-1.3-contributor-free","name":"Muse Spark 1.3 Free","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for coding and agentic workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meta/muse-spark-1.3"},"gemini-3-flash":{"id":"gemini-3-flash","name":"Gemini 3 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"longcat-2.5-preview-free":{"id":"longcat-2.5-preview-free","name":"LongCat 2.5 Preview Free","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meituan/longcat-2.5-preview"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Legacy model retained for compatibility with older integrations","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.4,"output":2.5,"cache_read":0.4},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://opencode.ai/inference/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://opencode.ai/inference/google/v1beta"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"big-pickle":{"id":"big-pickle","name":"Big Pickle","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"big-pickle","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-10-17","last_updated":"2025-10-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":160000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5"},"kimi-k2":{"id":"kimi-k2","name":"Kimi K2","description":"Legacy model retained for compatibility with older integrations","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.4,"output":2.5,"cache_read":0.4},"canonical_model_id":"moonshotai/kimi-k2-0905"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":1.07,"output":8.5,"cache_read":0.107},"canonical_model_id":"openai/gpt-5-codex"},"exo-free":{"id":"exo-free","name":"Exo Free","description":"Free-tier preview reasoning model with text and image input and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0}},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"glm-5.3-flash-free":{"id":"glm-5.3-flash-free","name":"GLM-5.3-Flash Free","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"hy3-free":{"id":"hy3-free","name":"Hy3 Free","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":190000,"input":192000,"output":64000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"tencent/hy3"},"muse-spark-1.2-contributor-free":{"id":"muse-spark-1.2-contributor-free","name":"Muse Spark 1.2 Free","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"status":"deprecated","provider":{"npm":"@ai-sdk/openai","api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"meta/muse-spark-1.2"},"grok-code":{"id":"grok-code","name":"Grok Code Fast 1","description":"Legacy model retained for compatibility with older integrations","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"xai/grok-code-fast-1"},"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"ling-3.1-flash-free":{"id":"ling-3.1-flash-free","name":"Ling 3.1 Flash Free","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","provider":{"api":"https://opencode.ai/inference/openai/v1"},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"inclusionai/ling-3.1-flash"}}},"aiand":{"id":"aiand","env":["AIAND_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.aiand.com/v1","name":"ai&","doc":"https://docs.aiand.com/","models":{"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"google/gemma-4-31b-it"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.75,"output":3.5,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":12.5,"cache_read":0.5},"canonical_model_id":"moonshotai/kimi-k3"},"motif-technologies/motif-3":{"id":"motif-technologies/motif-3","name":"Motif 3","description":"Motif 3 is a large-scale, decoder-only Mixture-of-Experts (MoE) language model with 314 billion total parameters and 13.2 billion parameters activated per token.","family":"motif","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2,"cache_read":0.2},"canonical_model_id":"motif-technologies/motif-3"},"deepseek-ai/deepseek-v4.1-flash":{"id":"deepseek-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.3,"output":0.6,"cache_read":0.02},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/deepseek-v4-pro":{"id":"deepseek-ai/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1,"output":2.5,"cache_read":0.25},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/deepseek-v4-flash":{"id":"deepseek-ai/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.25,"cache_read":0.08},"canonical_model_id":"deepseek/deepseek-v4-flash"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.08},"canonical_model_id":"openai/gpt-oss-120b"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.32,"output":3.2,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":3,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-27b"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":4,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-5.3":{"id":"zai-org/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":4,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/glm-5.3-flash":{"id":"zai-org/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048550,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"agnes":{"id":"agnes","env":["AGNES_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://apihub.agnes-ai.com/v1","name":"Agnes AI","doc":"https://agnes-ai.com/doc","models":{"agnes-2.5-pro-alpha":{"id":"agnes-2.5-pro-alpha","name":"Agnes 2.5 Pro Alpha","description":"Paid reasoning model for advanced coding, scientific reasoning, long-context analysis, agentic workflows, and multimodal understanding.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0.45,"output":0.9,"cache_read":0.0038}},"agnes-2.0-flash":{"id":"agnes-2.0-flash","name":"Agnes 2.0 Flash","description":"Fast and efficient model for agent workflows, tool calling, coding, and image understanding.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-25","last_updated":"2026-05-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":65536},"status":"deprecated","cost":{"input":0,"output":0}},"agnes-2.5-flash":{"id":"agnes-2.5-flash","name":"Agnes 2.5 Flash","description":"Upgraded model with improved coding, agent workflows, tool calling, and multimodal understanding.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07","last_updated":"2026-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":65536},"cost":{"input":0,"output":0}},"agnes-3.0-flash":{"id":"agnes-3.0-flash","name":"Agnes 3.0 Flash","description":"A new-generation Agnes AI text model with stronger Agnes Code task execution, tool orchestration, and trustworthy delivery.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09","last_updated":"2026-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":65536},"cost":{"input":0,"output":0}}}},"ofox":{"id":"ofox","env":["OFOX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ofox.ai/v1","name":"Ofox","doc":"https://ofox.ai/docs","models":{"volcengine/doubao-seed-2.0-lite":{"id":"volcengine/doubao-seed-2.0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.13,"output":0.76,"cache_read":0.03,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"volcengine/doubao-seed-2.1-turbo":{"id":"volcengine/doubao-seed-2.1-turbo","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.3536,"output":1.7696,"cache_read":0.068,"cache_write":0.0019},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"volcengine/doubao-seed-2.0-mini":{"id":"volcengine/doubao-seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.06,"output":0.56,"cache_read":0.02,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"volcengine/doubao-seed-2.0-code":{"id":"volcengine/doubao-seed-2.0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.67,"output":3.36,"cache_read":0.14,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"volcengine/doubao-seed-evolving":{"id":"volcengine/doubao-seed-evolving","name":"Seed Evolving","description":"Rolling ByteDance Seed model for rapidly updated reasoning, coding, and agent capabilities","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.884,"output":4.42,"cache_read":0.177,"cache_write":0.0025},"canonical_model_id":"bytedance-seed/seed-evolving"},"volcengine/doubao-seed-2.1-pro":{"id":"volcengine/doubao-seed-2.1-pro","name":"Seed 2.1 Pro","description":"Flagship ByteDance Seed 2.1 model for complex multimodal reasoning, coding, and agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.7072,"output":3.536,"cache_read":0.1416,"cache_write":0.002},"canonical_model_id":"bytedance-seed/seed-2.1-pro"},"volcengine/doubao-seed-2.0-pro":{"id":"volcengine/doubao-seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.67,"output":3.36,"cache_read":0.14,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"volcengine/doubao-seed-character":{"id":"volcengine/doubao-seed-character","name":"Seed Character","description":"ByteDance Seed model optimized for character-driven dialogue and consistent conversational behavior","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.177,"output":0.884,"cache_read":0.024,"cache_write":0.0025},"canonical_model_id":"bytedance-seed/seed-character"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.4,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.7-flashx":{"id":"z-ai/glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.072,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":1,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.083,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.0415,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":4.5,"input_audio":2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.0415,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":1,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":4.5,"input_audio":1.25},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.0415,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"provider":{"npm":"@ai-sdk/google","api":"https://api.ofox.ai/gemini/v1beta"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":1,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-pro-0423":{"id":"deepseek/deepseek-v4-pro-0423","name":"DeepSeek V4 Pro 0423","description":"DeepSeek V4 Pro initial snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v4-pro-0423"},"deepseek/deepseek-v4-flash-0423":{"id":"deepseek/deepseek-v4-flash-0423","name":"DeepSeek V4 Flash 0423","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.19,"output":0.51,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.29,"output":0.43,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.05,"output":0.4,"cache_read":0.01},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.18},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.ofox.ai/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"cache_write":6.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5,"cache_write":6.25}},"canonical_model_id":"openai/gpt-5.6-sol"},"bailian/qwen-turbo":{"id":"bailian/qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.05,"output":0.09,"cache_read":0.0086},"canonical_model_id":"alibaba/qwen-turbo"},"bailian/qwen3.7-plus":{"id":"bailian/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"bailian/qwen-plus":{"id":"bailian/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.12,"output":0.29,"cache_read":0.023},"canonical_model_id":"alibaba/qwen-plus"},"bailian/qwen-max":{"id":"bailian/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.35,"output":1.38,"cache_read":0.069},"canonical_model_id":"alibaba/qwen-max"},"bailian/qwen3.6-27b":{"id":"bailian/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.6-27b"},"bailian/qwen3-coder-next":{"id":"bailian/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"bailian/qwen3.8-flash":{"id":"bailian/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"bailian/qwen3.6-plus":{"id":"bailian/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"bailian/qwen3.5-flash":{"id":"bailian/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"bailian/qwen3.5-plus":{"id":"bailian/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"bailian/qwen3.5-397b-a17b":{"id":"bailian/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.55,"output":3.5,"cache_read":0.55},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"bailian/qwen3.5-122b-a10b":{"id":"bailian/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":2.29,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"bailian/qwen-vl-max":{"id":"bailian/qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.23,"output":0.58,"cache_read":0.046},"canonical_model_id":"alibaba/qwen-vl-max"},"bailian/qwen3.6-flash":{"id":"bailian/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.31},"canonical_model_id":"alibaba/qwen3.6-flash"},"bailian/qwen3-max":{"id":"bailian/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.36,"output":1.43,"cache_read":0.072},"canonical_model_id":"alibaba/qwen3-max"},"bailian/qwen3.8-max-0902":{"id":"bailian/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"bailian/qwen3-coder-plus":{"id":"bailian/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.8,"output":9,"cache_read":0.2,"cache_write":1},"canonical_model_id":"alibaba/qwen3-coder-plus"},"bailian/qwen3.7-max":{"id":"bailian/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"bailian/qwen-flash":{"id":"bailian/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.022,"output":0.22,"cache_read":0.0043,"cache_write":0.027},"canonical_model_id":"alibaba/qwen-flash"},"bailian/qwen3.8-max":{"id":"bailian/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"bailian/qwen3.6-max-preview":{"id":"bailian/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":2.15,"output":12.86,"cache_read":0.2,"cache_write":1.17},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"bailian/qwen3.5-35b-a3b":{"id":"bailian/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.29,"output":1.83,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"bailian/qwen3.5-27b":{"id":"bailian/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.29,"output":2.05,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-27b"},"bailian/qwen3-coder-flash":{"id":"bailian/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":2.5,"cache_read":0.06,"cache_write":0.27},"canonical_model_id":"alibaba/qwen3-coder-flash"},"bailian/qwen3.8-27b":{"id":"bailian/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1131072,"output":131072},"cost":{"input":0.45,"output":3.2,"cache_read":0.05,"cache_write":0.5625},"canonical_model_id":"alibaba/qwen3.8-27b"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.ofox.ai/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m2.5-lightning":{"id":"minimax/minimax-m2.5-lightning","name":"MiniMax-M2.5 Lightning","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1-lightning":{"id":"minimax/minimax-m2.1-lightning","name":"MiniMax-M2.1 Lightning","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/m2-her":{"id":"minimax/m2-her","name":"MiniMax-M2 Her","description":"MiniMax M2 variant tuned for conversational and character-driven agent interactions","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":2048},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2-Her"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"qwen/qwen-turbo":{"id":"qwen/qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.043,"output":0.09,"cache_read":0.0086},"canonical_model_id":"alibaba/qwen-turbo"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1064000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.12,"output":0.29,"cache_read":0.023},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen-max":{"id":"qwen/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8000},"cost":{"input":0.35,"output":1.38,"cache_read":0.069},"canonical_model_id":"alibaba/qwen-max"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.43,"output":2.57},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.2,"output":1.5},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11,"output":0.39,"cache_read":0.011,"cache_write":0.14},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.55,"output":3.5,"cache_read":0.55},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":2.29,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen-vl-max":{"id":"qwen/qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8000},"cost":{"input":0.23,"output":0.58,"cache_read":0.023},"canonical_model_id":"alibaba/qwen-vl-max"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.31},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":0.36,"output":1.43,"cache_read":0.072},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.71,"output":5.14,"cache_read":0.17,"cache_write":2.14},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.8,"output":9,"cache_read":0.2,"cache_write":1},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1064000,"output":64000},"cost":{"input":1.71,"output":5.14,"cache_read":0.17,"cache_write":2.14},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen-flash":{"id":"qwen/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.022,"output":0.22,"cache_read":0.0043,"cache_write":0.027},"canonical_model_id":"alibaba/qwen-flash"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.71,"output":5.14,"cache_read":0.17,"cache_write":2.14},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.6-max-preview":{"id":"qwen/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":2.15,"output":12.86,"cache_read":0.2,"cache_write":1.17},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":1.83,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.29,"output":2.05,"cache_read":0.29},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":2.5,"cache_read":0.06,"cache_write":0.27},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1131072,"output":131072},"cost":{"input":0.5,"output":1.71,"cache_read":0.043,"cache_write":0.63},"canonical_model_id":"alibaba/qwen3.8-27b"},"x-ai/grok-4.1-fast":{"id":"x-ai/grok-4.1-fast","name":"Grok 4.1 Fast","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"cost":{"input":4,"output":12,"cache_read":0.4},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"}}},"scaleway":{"id":"scaleway","env":["SCALEWAY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.scaleway.ai/v1","name":"Scaleway","doc":"https://www.scaleway.com/en/docs/generative-apis/","models":{"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT-OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.168225,"output":0.6729},"canonical_model_id":"openai/gpt-oss-120b"},"mistral-medium-3.5-128b":{"id":"mistral-medium-3.5-128b","name":"Mistral Medium 3.5 128B","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":1.68225,"output":8.41125},"canonical_model_id":"mistral/mistral-medium-2604"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":16384},"cost":{"input":1.00935,"output":1.00935},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":2.0187,"output":6.16825},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"status":"beta","cost":{"input":0.672,"output":2.016,"reasoning":2.016,"cache_read":0.134},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"lightonocr-3-0.8b":{"id":"lightonocr-3-0.8b","name":"LightOnOCR-3 0.8B","description":"LightOnOCR-3-0.8B is the lightest model of LightOn's LightOnOCR-3 family, with the same layout and visual understanding features as its larger siblings.","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-08","last_updated":"2026-10-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":2.464,"output":0},"canonical_model_id":"lightonai/lightonocr-3-0.8b"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":260000,"output":16384},"cost":{"input":0.841125,"output":2.523375},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen3-embedding-8b":{"id":"qwen3-embedding-8b","type":"embedding","name":"Qwen3 Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0.11215,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"status":"beta","cost":{"input":0.468,"output":0.936,"reasoning":0.936,"cache_read":0.0936},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.6729,"output":4.0374},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"bge-multilingual-gemma2":{"id":"bge-multilingual-gemma2","type":"embedding","name":"BGE Multilingual Gemma2","description":"Multilingual Gemma 2 based embedding model for retrieval and semantic search","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-07-25","last_updated":"2024-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8191,"output":0},"cost":{"input":0.11215,"output":0},"canonical_model_id":"baai/bge-multilingual-gemma2"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"status":"beta","cost":{"input":2.01,"output":6.15},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"beta","cost":{"input":0.280375,"output":1.68225},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"status":"beta","cost":{"input":0.280375,"output":0.56075},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"whisper-large-v3":{"id":"whisper-large-v3","type":"transcription","name":"Whisper Large v3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2023-09","release_date":"2023-09-01","last_updated":"2026-03-17","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":8192}},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"beta","cost":{"input":0.684,"output":3.762,"reasoning":3.762,"cache_read":0.137},"canonical_model_id":"alibaba/qwen3.8-27b"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"Mistral Small 3.2 24B Instruct (2506)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.168225,"output":0.392525},"canonical_model_id":"mistral/mistral-small-2506"}}},"opper":{"id":"opper","env":["OPPER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.opper.ai/v3/compat","name":"Opper","doc":"https://opper.ai/models","models":{"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.46488,"output":2.44062},"canonical_model_id":"google/gemma-4-31b-it"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":1.1622,"output":4.88124},"canonical_model_id":"openai/gpt-oss-120b"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.62708,"output":5.811},"canonical_model_id":"zhipuai/glm-5.2"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"sonar":{"id":"sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.825,"output":4.125,"cache_read":0.0825},"canonical_model_id":"google/gemini-3.8-flash"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":1.78812,"output":3.57624},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.11622,"output":0.488124},"canonical_model_id":"openai/gpt-oss-20b"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.12,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"tier":{"type":"context","size":524288}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24}},"canonical_model_id":"minimax/MiniMax-M3"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":6,"cache_read":0.63},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.5811,"output":2.3244},"canonical_model_id":"alibaba/qwen3-coder-next"},"sonar-reasoning-pro":{"id":"sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15},"canonical_model_id":"moonshotai/kimi-k3"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.69732,"output":2.78928},"canonical_model_id":"minimax/MiniMax-M2.7"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.75,"output":4.6488,"cache_read":0.44},"canonical_model_id":"zhipuai/glm-5.3"},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.5811,"output":2.44062},"canonical_model_id":"mistral/mistral-small-2603"},"sonar-pro":{"id":"sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.25,"output":0.66},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":0.5,"cache_read":0.07},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":45,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"openai/gpt-5.6-sol"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.5811,"output":3},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"bee":{"id":"bee","env":["BEE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.bee.heossi.com/bee","name":"Bee by HEOSSI","doc":"https://bee.heossi.com/docs/sdks","models":{"bee-hive":{"id":"bee-hive","name":"Bee Hive 1.0","description":"Scalable Bee serving with quantum-reasoning routing and multi-base domain-adapter routing for shared knowledge workflows, document intelligence, and lower-cost cloud inference. Domain adapters activate per customer domain after evaluation.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":5,"output":20,"cache_read":0.5,"cache_write":6}},"bee-swarm":{"id":"bee-swarm","name":"Bee Swarm 1.0","description":"Premium reasoning, multi-model cloud execution, large-document analysis, and high-value technical work. Includes the evolution engine and stronger eval gates by default.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":false,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":25,"output":150,"cache_read":2.5,"cache_write":30}},"bee-comb":{"id":"bee-comb","name":"Bee Comb 2.0","description":"Stronger local/workstation AI for coding, architecture, security reasoning, larger RAG, and domain-specific workflows. The default workhorse for engineers and dev teams.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":16384},"cost":{"input":1.25,"output":5,"cache_read":0.125,"cache_write":1.5}},"bee-cell":{"id":"bee-cell","name":"Bee Cell 1.0","description":"Fast hosted AI for chat, coding help, document question answering, image understanding, and lightweight technical reasoning.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.36}},"bee-buzz":{"id":"bee-buzz","name":"Bee Buzz 1.0","description":"Mid-tier production model for tool-using agents, multi-step workflows, and builder/integration work. Bigger context than Comb, deeper planning, faster than Hive on routine tasks.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":24576},"cost":{"input":2.5,"output":10,"cache_read":0.25,"cache_write":3}},"bee-brood":{"id":"bee-brood","name":"Bee Brood 1.0","description":"Production multimodal reasoning for image and video analysis, tool-enabled workflows, and multi-step planning at controlled cost.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-10","last_updated":"2026-09-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":16384},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.72}}}},"tensorx":{"id":"tensorx","env":["TENSORX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tensorx.ai/v1","name":"TensorX","doc":"https://docs.tensorx.ai/","models":{"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4,"cache_read":0.3,"cache_write":1.5},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.5,"output":4.5,"cache_read":0.375},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.35,"cache_write":1.75},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"status":"deprecated","cost":{"input":1,"output":3.2,"cache_read":0.25,"cache_write":1.25},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":1.75,"output":4.5,"cache_read":0.44},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4,"cache_read":0.3,"cache_write":1.5},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-5.3-flash"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.5,"output":1.5,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"deprecated","cost":{"input":1.75,"output":3.5,"cache_read":0.4375,"cache_write":2.185},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":2,"output":4,"cache_read":0.5},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.25,"output":0.3,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1-0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":8192},"status":"deprecated","cost":{"input":0.66,"output":2.6,"cache_read":0.165,"cache_write":0.825},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"status":"deprecated","cost":{"input":0.3,"output":0.5,"cache_read":0.075,"cache_write":0.375},"canonical_model_id":"deepseek/deepseek-v3.2"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.25,"output":4.5,"cache_read":0.3125},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":2.8,"cache_read":0.125,"cache_write":0.625},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.75},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":1,"output":4,"cache_read":0.25,"cache_write":1.25},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.4,"output":2,"cache_read":0.1},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.075,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":0.2,"cache_read":0.0375,"cache_write":0.1875},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3.8-flash-next":{"id":"qwen/qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":2.5,"output":6,"cache_read":0.63},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.5,"output":3.5,"cache_read":0.125,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-235b-a22b-2507":{"id":"qwen/qwen3-235b-a22b-2507","name":"Qwen3 235B-A22B-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":262144},"status":"deprecated","cost":{"input":0.072,"output":0.464,"cache_read":0.018,"cache_write":0.09},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":2.4,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"iteracompute":{"id":"iteracompute","env":["ITERACOMPUTE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.iteracompute.com/v1","name":"IteraCompute","doc":"https://iteracompute.com/docs.html","models":{"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.82,"output":2.88,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.08,"output":0.28,"cache_read":0.023},"canonical_model_id":"zhipuai/glm-5.3-flash"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.17,"output":0.7,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.79,"output":2.38,"cache_read":0.026},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731 (legacy alias)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.17,"output":0.7,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.69,"output":3.29,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":999999},"cost":{"input":2.35,"output":11.76,"cache_read":0.23},"canonical_model_id":"moonshotai/kimi-k3"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":524288},"cost":{"input":0.3,"output":1.19,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.95,"output":5.7,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"novita-ai":{"id":"novita-ai","env":["NOVITA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.novita.ai/openai","name":"Novita AI","doc":"https://novita.ai/docs/guides/introduction","models":{"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":98304,"output":16384},"cost":{"input":0.119,"output":0.2},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"baichuan/baichuan-m2-32b":{"id":"baichuan/baichuan-m2-32b","name":"Baichuan M2 32B","description":"Open medical model trained for clinical reasoning and patient interaction with on, off, and auto thinking modes","family":"baichuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"temperature":true,"release_date":"2025-08-13","last_updated":"2025-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.07,"output":0.07},"canonical_model_id":"baichuan/baichuan-m2-32b"},"xiaomimimo/mimo-v2.6-pro":{"id":"xiaomimimo/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomimimo/mimo-v2.5":{"id":"xiaomimimo/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.168,"output":0.336,"cache_read":0.0034},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomimimo/mimo-v2.6-flash":{"id":"xiaomimimo/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomimimo/mimo-v2.5-pro":{"id":"xiaomimimo/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.522,"output":1.044,"cache_read":0.0043},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":12288,"output":12288},"cost":{"input":0.135,"output":0.4},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-4-scout-17b-16e-instruct":{"id":"meta-llama/llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.18,"output":0.59},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama-3.1-8B-Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.02,"output":0.05},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-3.2-1b-instruct":{"id":"meta-llama/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Compact instruction-tuned Llama model for multilingual dialogue and on-device assistants","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":32000},"cost":{"input":0.02,"output":0.02},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"meta-llama/llama-4-maverick-17b-128e-instruct-fp8":{"id":"meta-llama/llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"deepseek/deepseek-r1-0528-qwen3-8b":{"id":"deepseek/deepseek-r1-0528-qwen3-8b","name":"DeepSeek R1 0528 Qwen3 8B","description":"DeepSeek R1 0528 distilled into Qwen3 8B for mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"release_date":"2025-05-29","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.06,"output":0.09},"canonical_model_id":"deepseek/deepseek-r1-0528-qwen3-8b"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.6,"output":3.2,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek_v3":{"id":"deepseek/deepseek_v3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16000},"cost":{"input":0.89,"output":0.89},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.44,"output":1.32,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.7,"output":2.5,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-r1-turbo":{"id":"deepseek/deepseek-r1-turbo","name":"DeepSeek R1 (Turbo)","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16000},"cost":{"input":0.7,"output":2.5},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek/deepseek-ocr-2":{"id":"deepseek/deepseek-ocr-2","name":"DeepSeek OCR 2","description":"High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0.03},"canonical_model_id":"deepseek/deepseek-ocr-2"},"nousresearch/hermes-2-pro-llama-3-8b":{"id":"nousresearch/hermes-2-pro-llama-3-8b","name":"Hermes 2 Pro Llama 3 8B","description":"Llama 3 fine-tune optimized for function calling, structured JSON responses, and dialogue","family":"hermes","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-04-30","last_updated":"2024-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.14,"output":0.14},"canonical_model_id":"nousresearch/hermes-2-pro-llama-3-8b"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 Instruct 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.8,"output":3.4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-instruct":{"id":"moonshotai/kimi-k2-instruct","name":"Kimi K2 Instruct","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"baidu/ernie-4.5-21B-a3b":{"id":"baidu/ernie-4.5-21B-a3b","name":"ERNIE 4.5 21B A3B","description":"Open ERNIE text MoE with 21B total and 3B active parameters","family":"ernie","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":120000,"output":8000},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"baidu/ernie-4.5-21b-a3b"},"baidu/cobuddy":{"id":"baidu/cobuddy","name":"CoBuddy","description":"Baidu Qianfan model optimized for code generation and tool-using coding agents","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-05-06","last_updated":"2026-05-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":0.28,"output":1.13,"cache_read":0.07},"canonical_model_id":"baidu/cobuddy"},"apodex/apodex-1.1-mini":{"id":"apodex/apodex-1.1-mini","name":"Apodex 1.1 Mini","description":"Open-weight 35B MoE model for long-horizon research, forecasting, and tool-assisted work","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-24","last_updated":"2026-08-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"apodex/apodex-1.1-mini"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash-fin":{"id":"inclusionai/ling-3.0-flash-fin","name":"Ling 3.0 Flash Fin","description":"Finance-enhanced model for financial research, multi-step investment workflows, and long-horizon planning and execution","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-27","last_updated":"2026-09-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-fin"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"inclusionai/ling-3.1-flash"},"inclusionai/ling-3.0-flash-sante":{"id":"inclusionai/ling-3.0-flash-sante","name":"Ling 3.0 Flash Sante","description":"Health and medicine specialization of Ling 3.0 Flash for medical reasoning and evidence-based retrieval","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-sante"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.05,"output":0.25},"canonical_model_id":"openai/gpt-oss-120b"},"microsoft/wizardlm-2-8x22b":{"id":"microsoft/wizardlm-2-8x22b","name":"WizardLM 2 8x22B","description":"Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2024-04-15","last_updated":"2024-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65535,"output":8000},"cost":{"input":0.62,"output":0.62},"canonical_model_id":"microsoft/wizardlm-2-8x22b"},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax L2 13B","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":3200},"cost":{"input":0.09,"output":0.09},"canonical_model_id":"gryphe/mythomax-l2-13b"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":524288}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen-mt-plus":{"id":"qwen/qwen-mt-plus","name":"Qwen MT Plus","description":"Flagship Qwen translation model with terminology control and support for 92 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-07-24","last_updated":"2025-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":0.25,"output":0.75},"canonical_model_id":"alibaba/qwen-mt-plus"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3-embedding-8b":{"id":"qwen/qwen3-embedding-8b","type":"embedding","name":"Qwen3 Embedding 8B","description":"Multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0.07,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-reranker-8b":{"id":"qwen/qwen3-reranker-8b","type":"reranking","name":"Qwen3 Reranker 8B","description":"Multilingual query-document reranking model supporting more than 100 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"cost":{"input":0.05,"output":0},"canonical_model_id":"alibaba/qwen3-reranker-8b"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":3.75,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.42,"output":3,"cache_read":0.085},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3-omni-30b-a3b-thinking":{"id":"qwen/qwen3-omni-30b-a3b-thinking","name":"Qwen3 Omni 30B A3B Thinking","description":"Multimodal reasoning model accepting text, images, video, and audio with text output","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.25,"output":0.97,"input_audio":2.2},"canonical_model_id":"alibaba/qwen3-omni-30b-a3b-thinking"},"minimaxai/minimax-m1-80k":{"id":"minimaxai/minimax-m1-80k","name":"MiniMax M1 80K","description":"Open hybrid-attention MoE reasoning model with a million-token context and an 80K thinking budget","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":40000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"minimax/minimax-m1-80k"},"mindai/macaron-v1-tall":{"id":"mindai/macaron-v1-tall","name":"Macaron V1 Tall","description":"Qwen3.6-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2475,"output":1.43,"cache_read":0.044},"canonical_model_id":"mindlab/macaron-v1-tall"},"mindai/macaron-v1-venti":{"id":"mindai/macaron-v1-venti","name":"Macaron V1 Venti","description":"GLM-5.2-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.825,"output":2.475,"cache_read":0.165},"canonical_model_id":"mindlab/macaron-v1-venti"},"baai/bge-m3":{"id":"baai/bge-m3","type":"embedding","name":"BGE M3","description":"Multilingual embedding model combining dense, sparse, and multi-vector retrieval","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-30","last_updated":"2024-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":0},"cost":{"input":0.01,"output":0},"canonical_model_id":"baai/bge-m3"},"baai/bge-reranker-v2-m3":{"id":"baai/bge-reranker-v2-m3","type":"reranking","name":"BGE Reranker V2 M3","description":"Lightweight multilingual cross-encoder that scores query-document relevance","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-15","last_updated":"2024-03-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":0},"cost":{"input":0.01,"output":0},"canonical_model_id":"baai/bge-reranker-v2-m3"},"zai-org/glm-4.5v":{"id":"zai-org/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai-org/glm-4.6":{"id":"zai-org/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/glm-4.7-flash":{"id":"zai-org/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/glm-4.5-air":{"id":"zai-org/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/glm-4.7":{"id":"zai-org/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/autoglm-phone-9b-multilingual":{"id":"zai-org/autoglm-phone-9b-multilingual","name":"AutoGLM Phone 9B Multilingual","description":"Multilingual visual model for operating smartphone interfaces and mobile agents","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-12-10","last_updated":"2025-12-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.035,"output":0.138},"canonical_model_id":"zhipuai/autoglm-phone-9b-multilingual"},"zai-org/glm-5.1":{"id":"zai-org/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/glm-5":{"id":"zai-org/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/glm-5.3":{"id":"zai-org/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/glm-4.6v":{"id":"zai-org/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"zai-org/glm-5.3-flash":{"id":"zai-org/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"paddlepaddle/paddleocr-vl":{"id":"paddlepaddle/paddleocr-vl","name":"PaddleOCR VL","description":"Compact vision-language model for multilingual document parsing, OCR, tables, formulas, and charts","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.02,"output":0.02},"canonical_model_id":"paddlepaddle/paddleocr-vl"}}},"cerebras":{"id":"cerebras","env":["CEREBRAS_API_KEY"],"npm":"@ai-sdk/cerebras","name":"Cerebras","doc":"https://inference-docs.cerebras.ai/models/overview","models":{"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.35,"output":0.75,"cache_read":0.35},"canonical_model_id":"openai/gpt-oss-120b"},"qwen-3.8-27b":{"id":"qwen-3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.99,"output":1.49,"cache_read":0.99},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"stepfun-step-plan":{"id":"stepfun-step-plan","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.com/step_plan/v1","name":"StepFun Step Plan (China)","doc":"https://platform.stepfun.com/docs/zh/step-plan/integrations/reasoning-api","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.7-flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"canonical_model_id":"stepfun/step-3.5-flash"},"step-router-v1":{"id":"step-router-v1","name":"Step Router v1","description":"StepFun routing model that dispatches requests to the appropriate Step model.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":256000}}}},"umans-ai-coding-plan":{"id":"umans-ai-coding-plan","env":["UMANS_AI_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.code.umans.ai/v1","name":"Umans AI Coding Plan","doc":"https://app.umans.ai/offers/code/docs","models":{"umans-glm-5.3":{"id":"umans-glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"umans-deepseek-v4.1-flash-1017":{"id":"umans-deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"umans-qwen3.6-35b-a3b":{"id":"umans-qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"umans-deepseek-v4.1-flash":{"id":"umans-deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"umans-coder":{"id":"umans-coder","name":"Umans Coder","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-glm-5.3-flash":{"id":"umans-glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131071},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"umans-kimi-k3":{"id":"umans-kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k3"},"umans-flash":{"id":"umans-flash","name":"Umans Flash","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"umans-deepseek-v4-flash-0731":{"id":"umans-deepseek-v4-flash-0731","name":"DeepSeek V4 Flash","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393215},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"}}},"google-vertex":{"id":"google-vertex","env":["GOOGLE_VERTEX_PROJECT","GOOGLE_VERTEX_LOCATION","GOOGLE_APPLICATION_CREDENTIALS"],"npm":"@ai-sdk/google-vertex","name":"Vertex","doc":"https://cloud.google.com/vertex-ai/generative-ai/docs/models","models":{"gemini-embedding-001":{"id":"gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"cost":{"input":0.15,"output":0},"canonical_model_id":"google/gemini-embedding-001"},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":60,"cache_read":0.05},"canonical_model_id":"google/gemini-3.1-flash-image"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"claude-haiku-4-5@20251001":{"id":"claude-haiku-4-5@20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"beta","cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"beta","cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"claude-fable-5@default":{"id":"claude-fable-5@default","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"gemini-3.1-pro-preview-customtools":{"id":"gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"beta","cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"claude-opus-4-7@default":{"id":"claude-opus-4-7@default","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"claude-sonnet-5-5@default":{"id":"claude-sonnet-5-5@default","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-sonnet-4-5@20250929":{"id":"claude-sonnet-4-5@20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"gemini-3.1-flash-lite-image":{"id":"gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":30,"cache_read":0.025},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30},"canonical_model_id":"google/gemini-2.5-flash-image"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1}},"gemini-2.5-flash-tts":{"id":"gemini-2.5-flash-tts","type":"speech","name":"Gemini 2.5 Flash TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-09-30","last_updated":"2025-12-10","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":32768,"input":8192,"output":16384},"cost":{"input":0.5,"output":10},"canonical_model_id":"google/gemini-2.5-flash-tts"},"gemini-2.5-pro-tts":{"id":"gemini-2.5-pro-tts","type":"speech","name":"Gemini 2.5 Pro TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-pro","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-09-30","last_updated":"2025-12-10","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":32768,"input":8192,"output":16384},"cost":{"input":1,"output":20},"canonical_model_id":"google/gemini-2.5-pro-tts"},"claude-opus-4-5@20251101":{"id":"claude-opus-4-5@20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":120,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image"},"claude-haiku-5-5@default":{"id":"claude-haiku-5-5@default","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"claude-opus-5@default":{"id":"claude-opus-5@default","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-opus-4-6@default":{"id":"claude-opus-4-6@default","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"claude-opus-5-5@default":{"id":"claude-opus-5-5@default","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-1@20250805":{"id":"claude-opus-4-1@20250805","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"claude-opus-4-8@default":{"id":"claude-opus-4-8@default","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-sonnet-5@default":{"id":"claude-sonnet-5@default","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-sonnet-4@20250514":{"id":"claude-sonnet-4@20250514","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"mistral-small-2503":{"id":"mistral-small-2503","name":"Mistral Small 3.1 24B","description":"Efficient multimodal model for instruction following, coding, reasoning, and function calling","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"provider":{"npm":"@ai-sdk/mistral","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/publishers/mistralai/models"},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"claude-fable-5-1@default":{"id":"claude-fable-5-1@default","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4@20250514":{"id":"claude-opus-4@20250514","name":"Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"codestral-2":{"id":"codestral-2","name":"Codestral 2","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"provider":{"npm":"@ai-sdk/mistral","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/publishers/mistralai/models"},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-2508"},"gemini-nano-banana-2.1":{"id":"gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Image model for prompt-driven generation, multi-turn editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":1.5,"output":30,"cache_read":0.15},"canonical_model_id":"google/gemini-nano-banana-2.1"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"mistral-medium-3":{"id":"mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/mistral","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/publishers/mistralai/models"},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-2505"},"claude-sonnet-4-6@default":{"id":"claude-sonnet-4-6@default","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/google-vertex/anthropic"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"google/gemma-4-26b-a4b-it-maas":{"id":"google/gemma-4-26b-a4b-it-maas","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"status":"alpha","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"xai/grok-4.1-fast-reasoning":{"id":"xai/grok-4.1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":30000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"xai/grok-4.20-non-reasoning":{"id":"xai/grok-4.20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":500000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":500000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":30000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4.1-fast-non-reasoning":{"id":"xai/grok-4.1-fast-non-reasoning","name":"Grok 4.1 Fast","description":"Fast Grok model for responsive chat, tool-assisted work, and low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":30000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"xai/grok-4.20-reasoning":{"id":"xai/grok-4.20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"moonshotai/kimi-k2-thinking-maas":{"id":"moonshotai/kimi-k2-thinking-maas","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.6,"output":2.5,"cache_read":0.06}},"meta/llama-4-scout-17b-16e-instruct-maas":{"id":"meta/llama-4-scout-17b-16e-instruct-maas","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1310720,"output":8192},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.25,"output":0.7},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta/llama-4-maverick-17b-128e-instruct-maas":{"id":"meta/llama-4-maverick-17b-128e-instruct-maas","name":"Llama 4 Maverick 17B 128E Instruct","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":8192},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.35,"output":1.15}},"meta/llama-3.3-70b-instruct-maas":{"id":"meta/llama-3.3-70b-instruct-maas","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.72,"output":0.72}},"deepseek-ai/deepseek-v3.1-maas":{"id":"deepseek-ai/deepseek-v3.1-maas","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.6,"output":1.7,"cache_read":0.06}},"deepseek-ai/deepseek-v3.2-maas":{"id":"deepseek-ai/deepseek-v3.2-maas","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-12-17","last_updated":"2026-04-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.56,"output":1.68,"cache_read":0.056}},"openai/gpt-oss-20b-maas":{"id":"openai/gpt-oss-20b-maas","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.07,"output":0.25,"cache_read":0.007}},"openai/gpt-oss-120b-maas":{"id":"openai/gpt-oss-120b-maas","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.09,"output":0.36},"canonical_model_id":"openai/gpt-oss-120b"},"qwen/qwen3-235b-a22b-instruct-2507-maas":{"id":"qwen/qwen3-235b-a22b-instruct-2507-maas","name":"Qwen3 235B A22B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-13","last_updated":"2025-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.22,"output":0.88}},"zai-org/glm-5-maas":{"id":"zai-org/glm-5-maas","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1,"output":3.2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"zai-org/glm-4.7-maas":{"id":"zai-org/glm-4.7-maas","name":"GLM-4.7","description":"GLM model for coding, tool use, and complex reasoning","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-06","last_updated":"2026-01-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":0.6,"output":2.2,"cache_read":0.06}},"zai-org/glm-5.2-maas":{"id":"zai-org/glm-5.2-maas","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${GOOGLE_VERTEX_ENDPOINT}/v1/projects/${GOOGLE_VERTEX_PROJECT}/locations/${GOOGLE_VERTEX_LOCATION}/endpoints/openapi"},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"}}},"nebius":{"id":"nebius","env":["NEBIUS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.tokenfactory.nebius.com/v1","name":"Nebius Token Factory","doc":"https://docs.tokenfactory.nebius.com/","models":{"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma-3-27b-it","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":110000,"input":100000,"output":8192},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"google/gemma-3-27b-it"},"nvidia/Nemotron-3-Ultra-550b-a55b":{"id":"nvidia/Nemotron-3-Ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":3,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/Nemotron-3.5-Super-VL":{"id":"nvidia/Nemotron-3.5-Super-VL","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for coding, visual understanding, and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron-3-Super-120B-A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/Nemotron-3_5-Lightning":{"id":"nvidia/Nemotron-3_5-Lightning","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.06,"output":0.24,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"reflection/beam":{"id":"reflection/beam","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.25,"output":1,"cache_read":0.05},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8000},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":8000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.2,"output":0.6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen3-30B-A3B-Instruct-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":8192},"cost":{"input":0.1,"output":0.3,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5-397B-A17B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":250000,"output":8192},"cost":{"input":0.6,"output":3.6,"cache_read":0.12},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3-Embedding-8B":{"id":"Qwen/Qwen3-Embedding-8B","type":"embedding","name":"Qwen3-Embedding-8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-10","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"input":40960,"output":0},"cost":{"input":0.01,"output":0},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"output":1048000},"cost":{"input":0.3,"output":1.2,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.75,"output":3.5,"cache_read":1.75},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":1024000},"cost":{"input":0.14,"output":0.28,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":979000,"output":979000},"cost":{"input":1.32,"output":3.96,"cache_read":0.04},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"gpt-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":124000,"output":8192},"cost":{"input":0.15,"output":0.6,"reasoning":0.6,"cache_read":0.15},"canonical_model_id":"openai/gpt-oss-120b"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":1024000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":1024000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"meta-models/muse-spark-open":{"id":"meta-models/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"NousResearch/Hermes-4-405B":{"id":"NousResearch/Hermes-4-405B","name":"Hermes-4-405B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":120000,"output":8192},"cost":{"input":1,"output":3,"reasoning":3,"cache_read":1},"canonical_model_id":"nousresearch/hermes-4-405b"}}},"llmtr":{"id":"llmtr","env":["LLMTR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://llmtr.com/v1","name":"LLMTR","doc":"https://llmtr.com/docs","models":{"muse-glimmer-30b-tr":{"id":"muse-glimmer-30b-tr","name":"Muse Glimmer 30B (TR)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":2,"output":5,"cache_read":0.5},"canonical_model_id":"meta/muse-glimmer-30b"},"trendyol-asure-12b":{"id":"trendyol-asure-12b","name":"Trendyol Asure 12B","description":"Turkish-language multimodal instruct model built on Gemma 3 12B for e-commerce text, chat, and image-text tasks","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-19","last_updated":"2026-02-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":40960},"cost":{"input":0.1,"output":0.5,"cache_read":0.025},"canonical_model_id":"trendyol/asure-12b"},"magibu-11b-v8":{"id":"magibu-11b-v8","name":"Magibu 11B v8","description":"Turkish-language chat model for instruction following and assistant flows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-06-05","last_updated":"2026-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.1,"output":0.5}},"gemma-4":{"id":"gemma-4","name":"Gemma 4","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":2,"output":5,"cache_read":0.5},"canonical_model_id":"google/gemma-4-31b-it"},"medgemma-4b":{"id":"medgemma-4b","name":"MedGemma 4B","description":"Multimodal medical-domain Gemma variant for text and image analysis","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-26","last_updated":"2026-08-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"status":"deprecated","cost":{"input":3,"output":5}},"qwen3-6-35b":{"id":"qwen3-6-35b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":5,"output":10},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.1},"canonical_model_id":"google/gemini-2.5-flash-lite"},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.58,"output":1.44,"cache_read":0.116},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.2"},"perplexity/sonar-deep-research":{"id":"perplexity/sonar-deep-research","name":"Sonar Deep Research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"},"upstage/solar-pro2":{"id":"upstage/solar-pro2","name":"Solar Pro 2","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro2"},"upstage/solar-pro3":{"id":"upstage/solar-pro3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.15,"output":0.6,"cache_read":0.03},"canonical_model_id":"upstage/solar-pro4"},"mimo/mimo-v2.5":{"id":"mimo/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.14,"output":0.28},"canonical_model_id":"xiaomi/mimo-v2.5"},"mimo/mimo-v2.5-pro":{"id":"mimo/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.435,"output":0.87},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"publicai/apertus-70b-instruct":{"id":"publicai/apertus-70b-instruct","name":"Apertus 70B Instruct","description":"Fully open 70B multilingual LLM supporting 1800+ languages with 65K context. Trained on 15T tokens of compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"cost":{"input":0.82,"output":2.92},"canonical_model_id":"swiss-ai/apertus-70b"},"publicai/apertus-8b-instruct":{"id":"publicai/apertus-8b-instruct","name":"Apertus 8B Instruct","description":"Fully open 8B multilingual LLM supporting 1800+ languages with 65K context. Trained on compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"swiss-ai/apertus-8b"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"mistral/voxtral-small-latest":{"id":"mistral/voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/voxtral-small-latest"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.2,"cache_read":0.08},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3-vl-plus":{"id":"qwen/qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"cost":{"input":0.2,"output":1.6,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3-vl-plus"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"status":"deprecated","cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":1,"output":5},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen-flash":{"id":"qwen/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.05,"output":0.4,"cache_read":0.01},"canonical_model_id":"alibaba/qwen-flash"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3-coder-flash"}}},"echo":{"id":"echo","env":["ECHO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://echo.tracerml.ai/v1","name":"Echo","doc":"https://echo.tracerml.ai/docs/api","models":{"echo":{"id":"echo","name":"Echo","description":"Adaptive model for coding, reasoning, and tool-driven agent workflows through one OpenAI-compatible endpoint","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"beta","cost":{"input":10,"output":50}}}},"tencent-tokenhub":{"id":"tencent-tokenhub","env":["TENCENT_TOKENHUB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://tokenhub.tencentmaas.com/v1","name":"Tencent TokenHub","doc":"https://cloud.tencent.com/document/product/1823/130050","models":{"hy3-preview":{"id":"hy3-preview","name":"Hy3 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"tencent/hy3-preview"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"tencent/hy3"}}},"salad-cloud":{"id":"salad-cloud","env":["SALAD_CLOUD_API_KEY"],"npm":"@saladtechnologies-oss/ai-sdk-provider","name":"SaladCloud AI Gateway","doc":"https://docs.salad.com/ai-gateway/explanation/overview","models":{"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Qwen MoE for agentic tasks, complex reasoning, code generation, and instruction following","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.09,"output":0.6},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"}}},"kilo":{"id":"kilo","env":["KILO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.kilo.ai/api/gateway","name":"Kilo Gateway","doc":"https://kilo.ai","models":{"mistralai/mistral-large-2407":{"id":"mistralai/mistral-large-2407","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large-2407"},"mistralai/mistral-large-4-0":{"id":"mistralai/mistral-large-4-0","name":"Mistral: Mistral Large 4","description":"Mistral Large 4 is a frontier multimodal (text and image input) model from Mistral AI built for reasoning, coding, and agentic workloads. It offers a 512K-token context window with up...","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2","description":"Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window. Devstral 2 supports exploring...","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mixtral-8x22b-instruct":{"id":"mistralai/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 25.08","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":204800},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"mistralai/voxtral-small-24b-2507":{"id":"mistralai/voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":26214},"cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Mistral Small 3.2 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.3,"cache_read":0.01},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Mistral Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":102400},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":102400},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-8b-2512"},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":26214},"cost":{"input":0.2,"output":0.6,"cache_read":0.02},"canonical_model_id":"mistral/mistral-saba"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Mistral Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-2603":{"id":"mistralai/mistral-small-2603","name":"Mistral Small 4","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mistral-nemo":{"id":"mistralai/mistral-nemo","name":"Mistral Nemo","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Mistral Large 3 2512 is Mistral’s most capable model to date, featuring a sparse mixture-of-experts architecture with 41B active parameters (675B total), and released under the Apache 2.0 license.","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistralai/mistral-medium-3-5":{"id":"mistralai/mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3b-2512"},"recursive/flint-v1.0:free":{"id":"recursive/flint-v1.0:free","name":"Recursive: Flint (free)","description":"Flint is an open-weight mixture-of-experts model from Recursive, with about 2.5B active parameters out of 40B total, built by Recursive's autoresearch system Eureka Machine for coding agents. It is suited...","family":"flint","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Fireworks: Ember-1","description":"Ember-1 is a specialized reasoning model from Fireworks Research, built on [Kimi K3](https://openrouter.ai/moonshotai/kimi-k3). It is designed to make every token go further: it produces shorter reasoning traces, using roughly 40%...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM-4.5-Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"GLM 5.2 is a large-scale reasoning model from Z.ai. It supports text input and output with a 1M-token context window, and is suited for long-horizon agent workflows, project-level software engineering,...","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.54,"output":1.98,"cache_read":0.099},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.3-prime":{"id":"z-ai/glm-5.3-prime","name":"Z.ai: GLM 5.3 Prime","description":"GLM-5.3-Prime is the high-speed variant of Z.ai's GLM-5.3, inheriting its full capabilities while delivering 1.5–2× the output throughput through inference acceleration. It supports text input and output with a 1M-token...","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":8.8,"cache_read":0.56}},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM-4.5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"Z.ai: GLM 5.3 FlashX","description":"GLM-5.3-FlashX is the high-speed variant of Z.ai's GLM-5.3-Flash, a native multimodal model delivering inference speeds of up to 200 tokens/s. Built on the same hybrid sparse and linear attention architecture...","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.09}},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"GLM-5.3 is a large-scale reasoning model from Z.ai, built for complex software engineering and long-horizon agent tasks. It supports text input and output with a 1M-token context window, and improves...","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM-5-Turbo","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"GLM-5.3-Flash is a native multimodal model from Z.ai. It is suited for efficient coding and long-horizon agent tasks. Its hybrid sparse and linear attention architecture maintains accurate long-context behavior while...","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":943717},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/lyria-3-pro-preview":{"id":"google/lyria-3-pro-preview","name":"Lyria 3 Pro Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-pro-preview"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Gemini 3.1 Flash Image, a.k.a. \"Nano Banana 2,\" is Google’s latest state of the art image generation and editing model, delivering Pro-level visual quality at Flash speed. It combines advanced...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.1,"output":0.34,"cache_read":0.1},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini 4 Argon is Google's flagship reasoning model, built for complex software engineering, deep mathematical and scientific problem solving, and autonomous multi-step agents. Google trained it to keep context across...","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":1.5,"cache_read":0.025,"cache_write":0.041667},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.04,"output":0.12},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":4.5,"reasoning":4.5,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Gemini 3.8 Flash is Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows.","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-2.5-pro-preview":{"id":"google/gemini-2.5-pro-preview","name":"Gemini 2.5 Pro Preview 06-05","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["pdf","image","text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.625,"output":5,"reasoning":5,"cache_read":0.0625,"cache_write":0.1875},"canonical_model_id":"google/gemini-2.5-pro-preview"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1,"output":6,"reasoning":6,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"Gemini 3.7 Flash is a multimodal model from Google for fast agentic workflows, coding, and complex multi-step reasoning. It is designed for tasks that require responsive performance and reliable multi-step...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.7-flash"},"google/lyria-3-clip-preview":{"id":"google/lyria-3-clip-preview","name":"Lyria 3 Clip Preview","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"google/lyria-3-clip-preview"},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) is Google's fastest, most cost-efficient Gemini image model, built for high-velocity developer pipelines and rapid-fire visual exploration. It delivers text-to-image generation...","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.15,"output":1.25,"reasoning":1.25,"cache_read":0.015,"cache_write":0.041667},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro is Google’s most advanced image-generation and editing model, built on Gemini 3 Pro. It extends the original Nano Banana with significantly improved multimodal reasoning, real-world grounding, and...","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"reasoning":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.125,"output":0.75,"reasoning":0.75,"cache_read":0.0125,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.04,"output":0.05,"cache_read":0.02},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.625,"output":5,"reasoning":5,"cache_read":0.0625,"cache_write":0.1875},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Gemini 3.6 Flash is a high-efficiency model from Google for coding, agentic workflows, and web and app development. It is designed to produce polished outputs with fewer unnecessary edits and...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.375,"output":1.875,"reasoning":1.875,"cache_read":0.0375,"cache_write":0.020833},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":1,"output":6,"reasoning":6,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.042,"output":0.22,"cache_read":0.021},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Gemini 3.5 Flash Lite is a high-efficiency model from Google with upgraded agentic capabilities. It is suited for subagents that execute focused tasks within complex, multi-agent workflows.","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.15,"output":1.25,"reasoning":1.25,"cache_read":0.015,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Nano Banana 2.1 (Gemini Nano Banana 2.1) is Google's image generation and editing model on the Flash tier, succeeding Nano Banana 2 and Nano Banana Pro. It improves product recontextualization,...","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["image","text","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.125,"output":0.75,"reasoning":0.75,"cache_read":0.0125,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"status":"deprecated","cost":{"input":0.05,"output":0.2,"reasoning":0.2,"cache_read":0.005,"cache_write":0.041667},"canonical_model_id":"google/gemini-2.5-flash-lite"},"google/gemma-2-27b-it":{"id":"google/gemma-2-27b-it","name":"Gemma 2 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-07-13","last_updated":"2024-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"cost":{"input":0.65,"output":0.65},"canonical_model_id":"google/gemma-2-27b-it"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":58982},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"nvidia/nemotron-3.5-super-vl:free":{"id":"nvidia/nemotron-3.5-super-vl:free","name":"NVIDIA: Nemotron 3.5 Super VL (free)","description":"NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. It takes text, images, documents, GUI screenshots, and video... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-3-ultra-550b-a55b:free":{"id":"nvidia/nemotron-3-ultra-550b-a55b:free","name":"NVIDIA: Nemotron 3 Ultra (free)","description":"NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3.5-content-safety:free":{"id":"nvidia/nemotron-3.5-content-safety:free","name":"NVIDIA: Nemotron 3.5 Content Safety (free)","description":"NVIDIA Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model from NVIDIA, fine-tuned from Google Gemma-3-4B. It moderates both inputs to and responses from LLMs and VLMs, accepting... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"NVIDIA Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model from NVIDIA, fine-tuned from Google Gemma-3-4B. It moderates both inputs to and responses from LLMs and VLMs, accepting...","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3.5-lightning:free":{"id":"nvidia/nemotron-3.5-lightning:free","name":"NVIDIA: Nemotron 3.5 Lightning (free)","description":"NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. It is suited for high-throughput agentic workloads and specialized tasks that... **Terms of service** For NVIDIA free endpoints (Super/Ultra/etc): Trial use only - do not submit personal or confidential data. Your use is logged for security purposes and to improve NVIDIA products and services. The logged session data for improvement purposes is not linked to your identity or any persistent identifier. For more information about our data processing practices, see our [Privacy Policy](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/). By interacting with this endpoint, you consent to our collection, recording, and use of such information and the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free":{"id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free","name":"NVIDIA: Nemotron 3 Nano Omni (free)","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.08,"output":0.45},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":182520},"cost":{"input":0.4,"output":1.76,"cache_read":0.08},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-super-120b-a12b:free":{"id":"nvidia/nemotron-3-super-120b-a12b:free","name":"NVIDIA: Nemotron 3 Super (free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nemotron 3.5 Lightning 30B A3B","description":"NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. It is suited for high-throughput agentic workloads and specialized tasks that...","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.025,"output":0.12,"cache_read":0.0125},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"NVIDIA: Nemotron 3.5 Super VL","description":"NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. It takes text, images, documents, GUI screenshots, and video...","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.06,"output":0.35,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"inception/mercury-2":{"id":"inception/mercury-2","name":"Inception: Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Inception: Mercury 2.5","description":"Mercury 2.5 is the fastest reasoning LLM, and the latest diffusion LLM (dLLM) from Inception. Instead of generating tokens sequentially, Mercury 2.5 produces and refines multiple tokens in parallel, achieving...","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.2,"output":0.75,"cache_read":0.02},"canonical_model_id":"inception/mercury-2.5"},"inception/mercury-3":{"id":"inception/mercury-3","name":"Inception: Mercury 3","description":"Mercury 3 is a diffusion-based language model from Inception, and the lab's most capable model to date. It is suited for agentic workflows and real-time applications where latency compounds, such...","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2,"output":0.7,"cache_read":0.02}},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"MiMo-V2.6-Pro is the flagship foundation model developed by Xiaomi. Built at a scale of over 1T parameters, it is designed to push the ceiling of capability for the most demanding...","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.003,"tiers":[{"input":0.8,"output":4,"cache_read":0.16,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.8,"output":4,"cache_read":0.16}},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo-V2.6-Flash is an open-source foundation model developed by Xiaomi. Built on a Mixture-of-Experts architecture with 309B total parameters and 15B activated per token, it employs a hybrid attention mechanism for...","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo-V2.6-Pro-UltraSpeed is the fast speed edition of Xiaomi's flagship foundation model, MiMo-V2.6-Pro. Built from the same 1T MiMo-V2.6-Pro checkpoint, it matches the original model in quality while delivering roughly 10x...","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":945000},"status":"deprecated","cost":{"input":0.435,"output":0.87,"cache_read":0.004,"tiers":[{"input":2,"output":6,"cache_read":0.4,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.4}},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"xingchen-agi/xing4.0-29b-a4b:free":{"id":"xingchen-agi/xing4.0-29b-a4b:free","name":"XingChen: Xing4.0 29B A4B (free)","description":"Xing4.0-29B-A4B is an open-weight mixture-of-experts language model from China Telecom's AI team (TeleAI) and the successor to the TeleChat series. It has 29B total parameters with 4B active per token...","family":"agi","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0,"output":0}},"elorian/glimpse":{"id":"elorian/glimpse","name":"Elorian: Glimpse","description":"Glimpse is a multimodal visual reasoning model from Elorian, built for visual understanding and complex visual reasoning in the physical world. It reasons directly over images and video rather than...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":256000},"cost":{"input":0.6,"output":3,"cache_read":0.06}},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Step 3.7 Flash is StepFun's latest high-efficiency multimodal Mixture-of-Experts model. It pairs a 196B-parameter language backbone with a vision encoder for native image and video understanding, activating roughly 11B parameters...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":230400},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"Step 5 Preview is StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters). It performs strongly in software engineering and professional...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"meituan/longcat-2.0":{"id":"meituan/longcat-2.0","name":"Meituan: LongCat 2.0","description":"LongCat 2.0 is a sparse mixture-of-experts language model from Meituan, with 48B active parameters out of 1.6T total. It is suited for coding, repository-level changes, long-horizon problem solving, and agentic...","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048756,"output":262144},"cost":{"input":0.75,"output":3,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"perceptron/perceptron-mk1":{"id":"perceptron/perceptron-mk1","name":"Perceptron: Perceptron Mk1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"perceptron/perceptron-mk1"},"perceptron/perceptron-mk1.5":{"id":"perceptron/perceptron-mk1.5","name":"Perceptron: Perceptron Mk1.5","description":"Perceptron Mk1.5 is Perceptron's embodied reasoning model for physical agents. It accepts text, image, video, and audio input, and answers with text plus optional structured annotations: points, boxes, polygons, tracks,...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":36864,"output":8192},"cost":{"input":0.15,"output":1.5}},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of...","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total. It is designed for general-purpose reasoning, coding, agentic and tool-use systems,...","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":471859},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"mancer/weaver":{"id":"mancer/weaver","name":"Mancer: Weaver (alpha)","description":"An attempt to recreate Claude-style verbosity, but don't expect the same level of coherence or memory. Meant for use in roleplay/narrative situations.","family":"alpha","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-08-02","last_updated":"2023-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":6000},"cost":{"input":0.4,"output":0.75},"canonical_model_id":"mancer/weaver"},"unbiased/pareto-26.9":{"id":"unbiased/pareto-26.9","name":"Pareto 26.9","description":"Pareto is a multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of general-purpose tasks.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25}},"unbiased/pareto-26.10":{"id":"unbiased/pareto-26.10","name":"Pareto 26.10","description":"Pareto is a multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of general-purpose tasks.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"relace/relace-apply-3":{"id":"relace/relace-apply-3","name":"Relace: Relace Apply 3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.85,"output":1.25},"canonical_model_id":"relace/relace-apply-3"},"relace/relace-search":{"id":"relace/relace-search","name":"Relace: Relace Search","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":1,"output":3},"canonical_model_id":"relace/relace-search"},"~x-ai/grok-latest":{"id":"~x-ai/grok-latest","name":"xAI: Grok Latest","description":"This model always redirects to the latest Grok model from xAI.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5}},"~moonshotai/kimi-latest":{"id":"~moonshotai/kimi-latest","name":"MoonshotAI: Kimi Latest","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.19,"output":8.94,"cache_read":0.19}},"reflection/beam-free":{"id":"reflection/beam-free","name":"Reflection: Beam 1.0 (free)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. Beam is Reflection's reasoning model for coding and agentic workflows, combining strong coding capabilities with fast, low-latency inference, served through Vercel AI Gateway at no cost. It supports function calling, structured outputs, and streaming, and takes text input only. Zero Data Retention is not available for this model.","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0,"reasoning":0,"cache_read":0}},"reflection/beam-1.0-501b-a23b":{"id":"reflection/beam-1.0-501b-a23b","name":"Reflection: Beam 1.0 501B A23B","description":"Beam 1.0 is Reflection's open-weight reasoning model for coding and long-horizon agentic workflows. It uses a sparse mixture-of-experts architecture with 501B total and 23B active parameters per token, and a...","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.24,"output":0.98,"cache_read":0.035},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"tencent/hunyuan-a13b-instruct":{"id":"tencent/hunyuan-a13b-instruct","name":"Tencent: Hunyuan A13B Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":117964},"cost":{"input":0.14,"output":0.57}},"tencent/hy-mt2-7b":{"id":"tencent/hy-mt2-7b","name":"Hy-MT2-7B","description":"Hy-MT2-7B is a 7B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided translation.","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-7b"},"tencent/hy3-preview":{"id":"tencent/hy3-preview","name":"Hy3 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.18,"output":0.6,"cache_read":0.06},"canonical_model_id":"tencent/hy3-preview"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"Tencent: Hy4 preview is a mixture-of-experts model from Tencent, with 49B active parameters out of 770B total. It is designed for coding agents, complex tool-use workflows, and productivity tasks that...","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy-mt2-30b-a3b":{"id":"tencent/hy-mt2-30b-a3b","name":"Hy-MT2-30B-A3B","description":"Hy-MT2-30B-A3B is Tencent's flagship translation model in the Hy-MT2 family. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and...","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295},"canonical_model_id":"tencent/hy-mt2-30b-a3b"},"tencent/hy-mt2-1.8b":{"id":"tencent/hy-mt2-1.8b","name":"Hy-MT2-1.8B","description":"Hy-MT2-1.8B is a compact 1.8B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided...","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.044,"output":0.177},"canonical_model_id":"tencent/hy-mt2-1.8b"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort:...","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":80000},"cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-4-maverick":{"id":"meta-llama/llama-4-maverick","name":"Meta: Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1875,"output":0.6525,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama-3.1-8B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.02,"output":0.04},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-3.2-1b-instruct":{"id":"meta-llama/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":54000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b-instruct"},"meta-llama/llama-4-scout":{"id":"meta-llama/llama-4-scout","name":"Meta: Llama 4 Scout","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"output":16384},"cost":{"input":0.08,"output":0.24},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.1-70b-instruct":{"id":"meta-llama/llama-3.1-70b-instruct","name":"Llama-3.1-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"meta-llama/llama-3.2-3b-instruct":{"id":"meta-llama/llama-3.2-3b-instruct","name":"Meta: Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":72000},"cost":{"input":0.05,"output":0.33},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"meta-llama/llama-guard-4-12b":{"id":"meta-llama/llama-guard-4-12b","name":"Meta: Llama Guard 4 12B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":16384},"cost":{"input":0.144,"output":0.144},"canonical_model_id":"meta/llama-guard-4-12b"},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek: DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":0.41}},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the cost-efficient tier of the V4.1 family. DeepSeek reports that it exceeds V4 Pro on performance, speed, and task...","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v3.1-terminus":{"id":"deepseek/deepseek-v3.1-terminus","name":"DeepSeek: DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163840,"output":147456},"cost":{"input":0.27,"output":1}},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"DeepSeek V4.1 Flash 1017 is the October 17, 2026 snapshot of DeepSeek V4.1 Flash, served by DeepSeek under the `deepseek-flash` API model name. V4.1 Flash is a sparse mixture-of-experts model...","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":384000},"cost":{"input":1.6,"output":3.2,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-chat-v3.1":{"id":"deepseek/deepseek-chat-v3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-chat-v3.1"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro.","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows.","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-chat-v3-0324":{"id":"deepseek/deepseek-chat-v3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":115200},"cost":{"input":0.29,"output":1.14,"cache_read":0.11},"canonical_model_id":"deepseek/deepseek-chat-v3-0324"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.2574,"output":1.0287},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048575,"output":943717},"cost":{"input":0.091,"output":0.182,"cache_read":0.0182},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of [DeepSeek V4 Flash 0731](https://openrouter.ai/deepseek/deepseek-v4-flash-0731) from DeepSeek, adding image understanding while matching the base model on text capabilities including agents,...","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.00686},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"nousresearch/hermes-3-llama-3.1-70b":{"id":"nousresearch/hermes-3-llama-3.1-70b","name":"Nous: Hermes 3 70B Instruct (retires Nov 2)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.7,"output":0.7},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"nousresearch/hermes-4-405b":{"id":"nousresearch/hermes-4-405b","name":"Nous: Hermes 4 405B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":1,"output":3},"canonical_model_id":"nousresearch/hermes-4-405b"},"nousresearch/hermes-3-llama-3.1-405b":{"id":"nousresearch/hermes-3-llama-3.1-405b","name":"Nous: Hermes 3 405B Instruct (retires Nov 2)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":1,"output":1},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-405b"},"rekaai/reka-flash-3":{"id":"rekaai/reka-flash-3","name":"Reka Flash 3","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"reka","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":58982},"cost":{"input":0.1,"output":0.2},"canonical_model_id":"rekaai/reka-flash-3"},"rekaai/reka-edge":{"id":"rekaai/reka-edge","name":"Reka Edge","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":14745},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"rekaai/reka-edge"},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 Instruct 0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"MoonshotAI: Kimi K2.7 Code is a coding-focused model in Moonshot AI's Kimi K2 family, built to complete end-to-end programming tasks reliably over long contexts. It uses a native multimodal mixture-of-experts...","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"MoonshotAI: Kimi K3.1 Pro","description":"Kimi K3.1 Pro is Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. It is built for long-horizon coding, 3D...","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.65,"output":3.41,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Meta: Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 Contributor is the cost-efficient contributor tier of Meta’s multimodal reasoning model for experimentation, learning, and early-stage agentic, multi-agent, and coding workflows. It is designed to track information...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer 30B is a dense, open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark and optimized for autonomous agents on consumer hardware. It is suited for long-horizon...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark 1.1 is a multimodal reasoning model from Meta, built for agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with a 1M-token context...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It is designed to keep track of information across extended tasks, work through...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Meta: Muse Spark 1.2 Contributor","description":"Muse Spark 1.2 contributor tier is a reasoning model from Meta designed for developers who want to start building at an even lower cost. It’s meaningfully cheaper than Muse Spark...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-08-21","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-wonder":{"id":"meta/muse-wonder","name":"Meta: Muse Wonder","description":"Muse Wonder is a frontier reasoning model from Meta, designed for complex personal and professional agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":4,"output":20,"cache_read":0.2}},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a reasoning model from Meta, designed for complex agentic tasks. It accepts text, images, video, audio, and PDF documents, returns text, and offers a 1M-token context...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is a coding-optimized open-weight model. Trained specifically for agentic coding workflows, it delivers higher first-attempt accuracy, reliable tool calling, and deep codebase understanding across the repos developers...","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"openrouter/bodybuilder":{"id":"openrouter/bodybuilder","name":"Body Builder (beta)","description":"Preview model for early access evaluation, prototyping, and compatibility testing","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-03-15","last_updated":"2026-03-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"status":"beta","cost":{"input":0,"output":0}},"openrouter/free":{"id":"openrouter/free","name":"OpenRouter Free Models Router","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-01","last_updated":"2026-02-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32768},"cost":{"input":0,"output":0}},"openrouter/auto":{"id":"openrouter/auto","name":"Auto Router","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-15","last_updated":"2026-03-15","modalities":{"input":["audio","image","pdf","text","video"],"output":["image","text"]},"open_weights":false,"limit":{"context":2000000,"output":32768},"cost":{"input":0,"output":0}},"openrouter/pareto-code":{"id":"openrouter/pareto-code","name":"Pareto Code Router","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-05-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":65536},"cost":{"input":0,"output":0}},"~anthropic/claude-fable-latest":{"id":"~anthropic/claude-fable-latest","name":"Anthropic: Claude Fable Latest ($$$$)","description":"This model always redirects to the latest model in the Claude Fable family.","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5}},"~anthropic/claude-sonnet-latest":{"id":"~anthropic/claude-sonnet-latest","name":"Anthropic: Claude Sonnet Latest","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"~anthropic/claude-haiku-latest":{"id":"~anthropic/claude-haiku-latest","name":"Anthropic: Claude Haiku Latest","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"~anthropic/claude-opus-latest":{"id":"~anthropic/claude-opus-latest","name":"Anthropic: Claude Opus Latest","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5}},"perplexity/sonar":{"id":"perplexity/sonar","name":"Perplexity: Sonar","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-27","last_updated":"2025-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127072,"output":114364},"cost":{"input":1,"output":1}},"perplexity/sonar-reasoning-pro":{"id":"perplexity/sonar-reasoning-pro","name":"Perplexity: Sonar Reasoning Pro","description":"Web-grounded reasoning model for multi-step research and cited answers","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8}},"perplexity/sonar-deep-research":{"id":"perplexity/sonar-deep-research","name":"Perplexity: Sonar Deep Research","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar-deep-research","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":115200},"cost":{"input":2,"output":8,"reasoning":3}},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Perplexity: Sonar Pro","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-07","last_updated":"2025-03-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"perplexity/sonar-pro-search":{"id":"perplexity/sonar-pro-search","name":"Perplexity: Sonar Pro Search","description":"Advanced Sonar search model for deeper research and cited synthesis","family":"sonar-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-30","last_updated":"2025-10-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8000},"cost":{"input":3,"output":15}},"kilo-auto/small":{"id":"kilo-auto/small","name":"Auto Small","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.4,"reasoning":0,"cache_read":0.005}},"kilo-auto/free":{"id":"kilo-auto/free","name":"Auto Free","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"cost":{"input":0,"output":0,"reasoning":0,"cache_read":0,"cache_write":0}},"kilo-auto/efficient":{"id":"kilo-auto/efficient","name":"Auto Efficient","description":"Routes each request to the cheapest model that gets the job done, based on continuously benchmarked accuracy and cost.","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"reasoning":0,"cache_read":0.0325,"cache_write":0.40625}},"kilo-auto/frontier":{"id":"kilo-auto/frontier","name":"Auto Frontier","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"reasoning":0,"cache_read":0.5,"cache_write":6.25}},"kilo-auto/balanced":{"id":"kilo-auto/balanced","name":"Auto Balanced","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"1970-01-01","last_updated":"1970-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"reasoning":0,"cache_read":0.0325,"cache_write":0.40625}},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Upstage: Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":117964},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-mini4":{"id":"upstage/solar-mini4","name":"Upstage: Solar Mini 4","description":"Solar Mini 4 is Upstage's compact, cost-efficient language model, a 35B-parameter mixture-of-experts with 3B active parameters and a 524K context window. It is built for agentic use cases where response...","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"upstage/solar-mini4"},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Upstage: Solar Pro 4","description":"Solar Pro 4 is a large language model from Upstage. It is suited for agentic workflows, office productivity, document-intensive work, and coding.","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"upstage/solar-pro4"},"stealth/glyph-cluster":{"id":"stealth/glyph-cluster","name":"Stealth: Glyph Cluster (free)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. Glyph Cluster is a reasoning model for coding and long-context analysis, served through Vercel AI Gateway during its stealth period at no cost. It supports function calling and streaming, takes text input only, and does not support structured outputs. Zero Data Retention is not available for this model.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0,"output":0,"reasoning":0,"cache_read":0}},"stealth/claude-opus-4.7":{"id":"stealth/claude-opus-4.7","name":"Stealth: Claude Opus 4.7 (20% off)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"reasoning":0,"cache_read":0.4,"cache_write":5}},"stealth/qwen3.6-plus":{"id":"stealth/qwen3.6-plus","name":"Stealth: Qwen3.6 Plus (50% off)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. This third-party-served variant of Qwen3.6 Plus is offered at 50% lower cost than standard Qwen3.6 Plus pricing and is not served by Alibaba or Kilo Code. Note: a surcharge applies to long-context workloads exceeding 256K input tokens.","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"reasoning":0,"cache_read":0.025,"cache_write":0.3125}},"stealth/claude-opus-4.6":{"id":"stealth/claude-opus-4.6","name":"Stealth: Claude Opus 4.6 (20% off)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"reasoning":0,"cache_read":0.4,"cache_write":5}},"stealth/claude-opus-4.8":{"id":"stealth/claude-opus-4.8","name":"Stealth: Claude Opus 4.8 (20% off)","description":"Your prompts and completions may be retained and used to train or improve the provider's services. This third-party-served variant of Claude Opus 4.8 is offered at 20% lower cost than standard Claude Opus 4.8 pricing and is not served by Anthropic or Kilo Code.","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"reasoning":0,"cache_read":0.4,"cache_write":5}},"stealth/claude-sonnet-4.6":{"id":"stealth/claude-sonnet-4.6","name":"Stealth: Claude Sonnet 4.6 (20% off)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.4,"output":12,"reasoning":0,"cache_read":0.24,"cache_write":3}},"dots-studio/dots-3-note-preview:free":{"id":"dots-studio/dots-3-note-preview:free","name":"Dots Studio: Dots3-Note Preview (free)","description":"Dots3-Note Preview is an open-weight mixture-of-experts model from Dots Studio, with 16B active parameters out of 280B total. It is the lightest model in the Dots 3 family and is...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":460800},"cost":{"input":0,"output":0}},"liquid/lfm-2.5-24b-a2b":{"id":"liquid/lfm-2.5-24b-a2b","name":"LiquidAI: LFM2.5-24B-A2B","description":"LFM2.5-24B-A2B is a mixture-of-experts reasoning model from Liquid AI with 24B total and 2.45B active parameters. It always reasons before answering and is built for agents that call tools. Liquid...","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"liquid/lfm-2.5-2.6b:free":{"id":"liquid/lfm-2.5-2.6b:free","name":"LiquidAI: LFM2.5-2.6B (free)","description":"LFM2.5-2.6B is a compact reasoning model from Liquid AI. It is suited for agent workflows, data extraction, RAG, and long-context processing. Liquid advises against using it for agentic coding or...","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192},"cost":{"input":0,"output":0}},"cognitivecomputations/dolphin-mistral-24b-venice-edition":{"id":"cognitivecomputations/dolphin-mistral-24b-venice-edition","name":"Venice: Uncensored","description":"Venice Uncensored Dolphin Mistral 24B Venice Edition is a fine-tuned variant of Mistral-Small-24B-Instruct-2501, developed by dphn.ai in collaboration with Venice.ai. This model is designed as an “uncensored” instruct-tuned LLM, preserving...","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.2,"output":0.9},"canonical_model_id":"cognitivecomputations/dolphin-mistral-24b-venice-edition"},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"IBM: Granite 4.2 8B","description":"Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning. It supports full, low-effort,...","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"ibm-granite/granite-4.0-h-micro":{"id":"ibm-granite/granite-4.0-h-micro","name":"IBM: Granite 4.0 Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"granite","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131000,"output":117900},"cost":{"input":0.017,"output":0.112}},"aion-labs/aion-3.0-mini":{"id":"aion-labs/aion-3.0-mini","name":"AionLabs: Aion-3.0-Mini","description":"Aion-3.0 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the DeepSeek family of models. It uses a collaborative generation process in which multiple specialized models each...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18},"canonical_model_id":"aion-labs/aion-3.0-mini"},"aion-labs/aion-3.5-mini":{"id":"aion-labs/aion-3.5-mini","name":"AionLabs: Aion 3.5 Mini","description":"Aion 3.5 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It is the smaller, lower-cost sibling of Aion 3.5 and uses...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18}},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"AionLabs: Aion-RP 1.0 (8B)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.8,"output":1.6},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"aion-labs/aion-3.0":{"id":"aion-labs/aion-3.0","name":"AionLabs: Aion-3.0","description":"Aion-3.0 is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It uses a collaborative generation process in which multiple specialized models each contribute...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75},"canonical_model_id":"aion-labs/aion-3.0"},"aion-labs/aion-2.0":{"id":"aion-labs/aion-2.0","name":"AionLabs: Aion-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-21","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.8,"output":1.6,"cache_read":0.2},"canonical_model_id":"aion-labs/aion-2.0"},"aion-labs/aion-3.5":{"id":"aion-labs/aion-3.5","name":"AionLabs: Aion 3.5","description":"Aion 3.5 is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It uses a collaborative generation process in which multiple specialized models each...","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75}},"moondream/moondream3.1-9b-a2b":{"id":"moondream/moondream3.1-9b-a2b","name":"Moondream: Moondream 3.1","description":"Moondream 3.1 is a vision-language model from Moondream, the successor to Moondream 3 Preview, with 9B total parameters and 2B active (mixture-of-experts). It handles visual question answering, captioning, OCR, and...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"moondream/moondream3-preview":{"id":"moondream/moondream3-preview","name":"Moondream: Moondream 3 Preview","description":"Moondream 3 Preview is a vision-language model from Moondream with 9B total parameters and 2B active (mixture-of-experts). It handles visual question answering, captioning, OCR, and visual reasoning over images, with...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"~deepseek/deepseek-flash-latest":{"id":"~deepseek/deepseek-flash-latest","name":"DeepSeek: DeepSeek Flash Latest","description":"This model always redirects to the latest model in the DeepSeek Flash family.","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.02,"output":1,"cache_read":0.015}},"~deepseek/deepseek-pro-latest":{"id":"~deepseek/deepseek-pro-latest","name":"DeepSeek: DeepSeek Pro Latest","description":"This model always redirects to the latest model in the DeepSeek Pro family.","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-14","last_updated":"2026-09-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":393216},"cost":{"input":0.1188,"output":8,"cache_read":0.12}},"~deepseek/deepseek-v4-flash-latest":{"id":"~deepseek/deepseek-v4-flash-latest","name":"DeepSeek: DeepSeek V4 Flash Latest","description":"This model always redirects to the latest model in the DeepSeek V4 Flash family.","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-01","last_updated":"2026-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":0.0422,"output":0.1694,"cache_read":0.028}},"amazon/nova-micro-v1":{"id":"amazon/nova-micro-v1","name":"Amazon: Nova Micro 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":5120},"cost":{"input":0.035,"output":0.14},"canonical_model_id":"amazon/nova-micro-v1"},"amazon/nova-premier-v1":{"id":"amazon/nova-premier-v1","name":"Amazon: Nova Premier 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-30","last_updated":"2025-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"status":"deprecated","cost":{"input":2.5,"output":12.5,"cache_read":0.625},"canonical_model_id":"amazon/nova-premier"},"amazon/nova-2-lite-v1":{"id":"amazon/nova-2-lite-v1","name":"Amazon: Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"amazon/nova-2-lite-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Amazon: Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.8,"output":3.2},"canonical_model_id":"amazon/nova-pro-v1"},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Amazon: Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"amazon/nova-lite-v1"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"inclusionAI: Ling 3.0 Flash VL","description":"Ling 3.0 Flash VL builds on Ling 3.0 Flash (124B total / 5.5B active MoE from InclusionAI), further strengthening its language capabilities while adding native visual perception and advanced visual...","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"inclusionAI: Ling 3.0 Flash","description":"*Ling-3.0-flash* is a *124B-parameter Mixture-of-Experts (MoE) model*, with approximately *5.1B parameters activated per token*. The model is designed with *token efficiency and production-scale agentic inference* as key priorities, enabling developers...","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"inclusionAI: Ling 3.1 Flash","description":"Ling 3.1 Flash is a hybrid reasoning mixture-of-experts model from inclusionAI, with 25B active parameters out of 560B total.","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"inclusionai/ling-3.1-flash"},"~google/gemini-pro-latest":{"id":"~google/gemini-pro-latest","name":"Google: Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["audio","pdf","image","text","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":2,"output":10,"reasoning":10,"cache_read":0.1,"cache_write":0.1875}},"~google/gemini-flash-latest":{"id":"~google/gemini-flash-latest","name":"Google: Gemini Flash Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"reasoning":3.75,"cache_read":0.075,"cache_write":0.041667}},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"output":5529},"cost":{"input":0.35,"output":0.65},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"~unbiased/pareto-latest":{"id":"~unbiased/pareto-latest","name":"Pareto Latest","description":"This model always redirects to the latest model in the Pareto family.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"openai/gpt-6-astra-pro":{"id":"openai/gpt-6-astra-pro","name":"OpenAI: GPT-6 Astra Pro ($$$$)","description":"GPT-6 Astra Pro is the same underlying model as [GPT-6 Astra](https://openrouter.ai/openai/gpt-6-astra), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.03,"output":0.15,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-4o-mini-2024-07-18":{"id":"openai/gpt-4o-mini-2024-07-18","name":"OpenAI: GPT-4o-mini (2024-07-18)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":150,"output":600},"canonical_model_id":"openai/o1-pro"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5-image-mini":{"id":"openai/gpt-5-image-mini","type":"image","name":"GPT-5 Image Mini","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["pdf","image","text"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":2,"cache_read":0.25},"canonical_model_id":"openai/gpt-5-image-mini"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon...","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5-image":{"id":"openai/gpt-5-image","type":"image","name":"GPT-5 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-10-01","release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":10,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-5-image"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT-5.6 Luna","description":"GPT-5.6 Luna Pro is the same underlying model as [GPT-5.6 Luna](https://openrouter.ai/openai/gpt-5.6-luna), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.018,"output":0.09,"cache_read":0.009},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-audio-mini":{"id":"openai/gpt-audio-mini","type":"chat","name":"GPT-Audio Mini","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio","pdf"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":2.4},"canonical_model_id":"openai/gpt-audio-mini"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6.1-sol-pro":{"id":"openai/gpt-6.1-sol-pro","name":"OpenAI: GPT-6.1 Sol Pro","description":"GPT-6.1 Sol Pro is the same underlying model as [GPT-6.1 Sol](https://openrouter.ai/openai/gpt-6.1-sol), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. **Cost note:** pro mode spends far more...","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"openai/gpt-audio":{"id":"openai/gpt-audio","type":"chat","name":"GPT-Audio","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","audio","pdf"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-audio"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"GPT-6 Luna is the fast, cost-efficient model in OpenAI's GPT-6 series, positioned below GPT-6 Sol. It is suited for high-volume and latency-sensitive workloads such as chat, classification, and lightweight agentic...","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. It is suited for demanding professional...","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-6-luna-pro":{"id":"openai/gpt-6-luna-pro","name":"OpenAI: GPT-6 Luna Pro","description":"GPT-6 Luna Pro is the same underlying model as [GPT-6 Luna](https://openrouter.ai/openai/gpt-6-luna), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/o4-mini-high":{"id":"openai/o4-mini-high","name":"OpenAI: o4 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275}},"openai/gpt-chat-latest":{"id":"openai/gpt-chat-latest","name":"OpenAI: GPT Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5}},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT-5.6 Terra","description":"GPT-5.6 Terra Pro is the same underlying model as [GPT-5.6 Terra](https://openrouter.ai/openai/gpt-5.6-terra), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"GPT-5.6 Terra is a balanced model in OpenAI's GPT-5.6 series, positioned between the flagship Sol tier and the cost-efficient Luna tier. It is suited for everyday coding, reasoning, and agentic...","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"GPT-6.1 Sol is an upgrade to GPT-6 Sol from OpenAI, positioned below the flagship GPT-6 Astra in the GPT-6 series. It is suited for agentic coding, computer use, document-heavy professional...","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.4-image-2":{"id":"openai/gpt-5.4-image-2","type":"image","name":"GPT-5.4 Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":8,"output":15,"cache_read":2},"canonical_model_id":"openai/gpt-5.4-image-2"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-3.5-turbo-16k":{"id":"openai/gpt-3.5-turbo-16k","name":"OpenAI: GPT-3.5 Turbo 16k","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2023-08-28","last_updated":"2023-08-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":3,"output":4}},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o3":{"id":"openai/o3","name":"o3","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":4096},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT-5.6 Sol","description":"GPT-5.6 Sol Pro is the same underlying model as [GPT-5.6 Sol](https://openrouter.ai/openai/gpt-5.6-sol), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"GPT-5.6 Luna is a fast, cost-efficient model in OpenAI's GPT-5.6 series. It is suited for high-volume, latency-sensitive tasks such as chat, classification, and lightweight agentic workflows, providing capable reasoning for...","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-6-sol-pro":{"id":"openai/gpt-6-sol-pro","name":"OpenAI: GPT-6 Sol Pro","description":"GPT-6 Sol Pro is the same underlying model as [GPT-6 Sol](https://openrouter.ai/openai/gpt-6-sol), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2-chat":{"id":"openai/gpt-5.2-chat","name":"OpenAI: GPT-5.2 Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-3.5-turbo-instruct":{"id":"openai/gpt-3.5-turbo-instruct","name":"OpenAI: GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-09-28","last_updated":"2023-09-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1.5,"output":2}},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","pdf","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-3.5-turbo-0613":{"id":"openai/gpt-3.5-turbo-0613","name":"OpenAI: GPT-3.5 Turbo (older v0613)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":4095,"output":3685},"cost":{"input":1,"output":2}},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"GPT-5.6 Sol is the flagship model in OpenAI's GPT-5.6 series. It is suited for complex reasoning, coding, and agentic workflows, and is particularly strong at command-line and multi-step coding tasks...","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"OpenAI: o3 Mini High","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-02-12","last_updated":"2025-02-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"prism-ml/ternary-bonsai-2-27b":{"id":"prism-ml/ternary-bonsai-2-27b","name":"PrismML: Ternary Bonsai 2 27B","description":"Bonsai 2 27B is a 27B-parameter reasoning model from PrismML derived from Qwen3.8-27B. It supports coding, mathematics, tool calling, and image understanding with a 262K-token context window. Ternary compression shrinks...","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.075,"output":0.5,"cache_read":0.0375}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Inference.net: Schematron V2 Turbo","description":"Schematron V2 Turbo is a 3B-parameter HTML-to-JSON extraction model from Inference.net. It prioritizes throughput for high-volume extraction workloads. Extraction instructions must be supplied through a JSON schema in response_format rather...","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Inference.net: Schematron V2 Small","description":"Schematron V2 Small is a 3B-parameter HTML-to-JSON extraction model from Inference.net. It prioritizes extraction quality for complex schemas and long pages. Extraction instructions must be supplied through a JSON schema...","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.05},"canonical_model_id":"inference-net/schematron-v2-small"},"microsoft/phi-4":{"id":"microsoft/phi-4","name":"Microsoft: Phi 4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-01-10","last_updated":"2025-01-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":14745},"cost":{"input":0.056,"output":0.112}},"poolside/laguna-s-2.1:free":{"id":"poolside/laguna-s-2.1:free","name":"Poolside: Laguna S 2.1 (free)","description":"Laguna S 2.1 is the latest coding agent model from [Poolside](). Laguna S 2.1 is a 118B total parameter model with 8B active parameters, scoring 70.2% on Terminal-Bench 2.1 and...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-xs-2.1:free":{"id":"poolside/laguna-xs-2.1:free","name":"Poolside: Laguna XS 2.1 (free)","description":"Laguna XS 2.1 is the latest coding agent model in the 33B-A3B category from [Poolside](https://poolside.ai/) and a step forward from their Laguna XS.2 model (released in April 2026). It combines...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Poolside: Laguna XS 2.1 (retires Oct 31)","description":"Laguna XS 2.1 is the latest coding agent model in the 33B-A3B category from [Poolside](https://poolside.ai/) and a step forward from their Laguna XS.2 model (released in April 2026). It combines...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.2,"cache_read":0.05},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Poolside: Laguna S 2.1 (retires Oct 31)","description":"Laguna S 2.1 is the latest coding agent model from [Poolside](). Laguna S 2.1 is a 118B total parameter model with 8B active parameters, scoring 70.2% on Terminal-Bench 2.1 and...","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.1,"output":0.2,"cache_read":0.01},"canonical_model_id":"poolside/laguna-s-2.1"},"~openai/gpt-sol-latest":{"id":"~openai/gpt-sol-latest","name":"OpenAI: GPT Sol Latest","description":"This model always redirects to the latest model in the OpenAI GPT Sol family.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"~openai/gpt-terra-latest":{"id":"~openai/gpt-terra-latest","name":"OpenAI: GPT Terra Latest","description":"This model always redirects to the latest model in the OpenAI GPT Terra family.","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5}},"~openai/gpt-luna-latest":{"id":"~openai/gpt-luna-latest","name":"OpenAI: GPT Luna Latest","description":"This model always redirects to the latest model in the OpenAI GPT Luna family.","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"~openai/gpt-mini-latest":{"id":"~openai/gpt-mini-latest","name":"OpenAI: GPT Mini Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075}},"~openai/gpt-astra-latest":{"id":"~openai/gpt-astra-latest","name":"OpenAI: GPT Astra Latest ($$$$)","description":"This model always redirects to the latest model in the OpenAI GPT Astra family.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["pdf","image","text"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"thedrummer/skyfall-36b-v2":{"id":"thedrummer/skyfall-36b-v2","name":"TheDrummer: Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.55,"output":0.8,"cache_read":0.25},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"thedrummer/cydonia-24b-v4.1":{"id":"thedrummer/cydonia-24b-v4.1","name":"TheDrummer: Cydonia 24B V4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.3,"output":0.5,"cache_read":0.15},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"thedrummer/unslopnemo-12b":{"id":"thedrummer/unslopnemo-12b","name":"TheDrummer: UnslopNemo 12B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-11-08","last_updated":"2024-11-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":819200},"cost":{"input":0.4,"output":0.4},"canonical_model_id":"thedrummer/unslopnemo-12b"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Sonnet 5 is Anthropic's most capable Sonnet-class model, with frontier performance across coding, agents, and professional work. It supports adaptive thinking with selectable reasoning effort levels (low, medium, high, max,...","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Claude Opus 5 is Anthropic’s flagship model for demanding reasoning, coding, and long-horizon agentic work. It is particularly strong at end-to-end software tasks, code review and bug finding, visual analysis...","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude Fable 5 is a Mythos-class model from Anthropic, built for autonomous knowledge work and coding. It supports text, image, and file inputs with text output, with reasoning support and...","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude Fable 5.1 improves on Claude Fable 5 across the board, with the biggest gains in agentic coding, long-running agentic workflows, and knowledge work: long code refactors, front-end and visual...","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work, succeeding Claude Opus 5. It is particularly strong at multi-step changes in large codebases, code...","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Claude Haiku 5.5 is Anthropic's small, fast model for high-volume, cost-sensitive work such as summarization, subagents, and browser use. It succeeds Claude Haiku 4.5 with stronger coding, computer use, and...","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Claude Sonnet 5.5 is Anthropic's Sonnet-class model for well-scoped everyday work, succeeding Claude Sonnet 5 as a direct upgrade. It is especially strong at building features, fixing bugs, and producing...","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Claude Opus 4.8 is Anthropic's most capable generally available model in the Opus family. It supports text, image, and file inputs with text output, with reasoning support and a 1M-token...","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":3686},"cost":{"input":0.08,"output":0.11},"canonical_model_id":"gryphe/mythomax-l2-13b"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Fugu Ultra v2 is the higher-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to...","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Fugu Max is the cost-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to route...","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/sakana-namazu":{"id":"sakana/sakana-namazu","name":"Sakana Namazu","description":"Sakana Namazu is a Japanese-specialized reasoning model from Sakana AI, based on Kimi K2.6 with additional training for Japanese language and business contexts. It is suited for Japanese instruction following,...","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Fugu Ultra is the higher-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system: a language model trained to route...","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"minimax/minimax-01":{"id":"minimax/minimax-01","name":"MiniMax: MiniMax-01","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000192,"output":900172},"cost":{"input":0.2,"output":1.1}},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax: MiniMax M3.1 Flash","description":"MiniMax-M3.1-Flash is a multimodal model from MiniMax. Compared with M3, it delivers significant improvements across multiple benchmarks, particularly in complex agentic tasks, coding, and cybersecurity. It is built for complex,...","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":524288},"cost":{"input":0.09,"output":0.5,"cache_read":0.01},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m1":{"id":"minimax/minimax-m1","name":"MiniMax: MiniMax M1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":40000},"cost":{"input":0.4,"output":2.2}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.224,"output":0.88,"cache_read":0.0448},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2-her":{"id":"minimax/minimax-m2-her","name":"MiniMax-M2 Her","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":2048},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2-Her"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.27,"output":0.95,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2"},"sao10k/l3-lunaris-8b":{"id":"sao10k/l3-lunaris-8b","name":"Sao10K: Llama 3 8B Lunaris","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-13","last_updated":"2024-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":7372},"cost":{"input":0.04,"output":0.05},"canonical_model_id":"sao10k/l3-lunaris-8b"},"sao10k/l3.1-euryale-70b":{"id":"sao10k/l3.1-euryale-70b","name":"Sao10K: Llama 3.1 Euryale 70B v2.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.85,"output":0.85},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"status":"deprecated","cost":{"input":0.17,"output":0.25},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.21,"output":1.9,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Qwen3.7-Plus is a cost-effective model in Alibaba's Qwen3.7 series. It supports text and image input with text output, building on the series' text capabilities with a comprehensive upgrade to its...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.32,"output":1.28,"cache_read":0.064,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3-coder":{"id":"qwen/qwen3-coder","name":"Qwen: Qwen3 Coder 480B A35B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":1,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen/qwen3.5-plus-02-15":{"id":"qwen/qwen3.5-plus-02-15","name":"Qwen: Qwen3.5 Plus 2026-02-15","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.26,"output":1.56}},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.26,"output":0.78,"cache_read":0.052,"cache_write":0.325},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3-coder-30b-a3b-instruct":{"id":"qwen/qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"qwen/qwen3.5-plus-20260420":{"id":"qwen/qwen3.5-plus-20260420","name":"Qwen: Qwen3.5 Plus 2026-04-20","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.8,"cache_write":0.375}},"qwen/qwen3.8-max-prime":{"id":"qwen/qwen3.8-max-prime","name":"Qwen 3.8 Max Prime","description":"Qwen3.8 Max Prime is a higher-throughput variant of Qwen3.8 Max from Alibaba's Qwen team, served as a separate SKU at a higher price point. It accepts text, image, and video...","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"qwen/qwen2.5-vl-72b-instruct":{"id":"qwen/qwen2.5-vl-72b-instruct","name":"Qwen: Qwen2.5 VL 72B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":115200},"cost":{"input":0.8,"output":1,"cache_read":0.4},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"qwen/qwen-2.5-72b-instruct":{"id":"qwen/qwen-2.5-72b-instruct","name":"Qwen2.5 72B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":16384},"cost":{"input":0.36,"output":0.4}},"qwen/qwen3.5-flash-02-23":{"id":"qwen/qwen3.5-flash-02-23","name":"Qwen: Qwen3.5-Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-25","last_updated":"2026-02-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.065,"output":0.26}},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.45,"output":2.7},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen: Qwen3 30B A3B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.09,"output":0.3},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":909000},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.12,"output":0.8,"cache_read":0.07},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen3.8 Flash is a multimodal reasoning model from Alibaba. It is suited for coding assistance, agentic workflows, visual understanding, document and codebase analysis, desktop interaction, chart analysis, and long-video analysis.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen3.8 Omni Flash is an omni-modal reasoning model from Alibaba, the first Qwen model built around agentic capabilities with native audio-video understanding. It is suited for audio-video analysis and summarization,...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.325,"output":1.95,"cache_write":0.40625},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.39,"output":2.34},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26,"output":2.08},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-14b":{"id":"qwen/qwen3-14b","name":"Qwen: Qwen3 14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.24}},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1875,"output":1.125,"cache_write":0.234375},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"Qwen3.8 Max 0902 is an updated snapshot of Qwen3.8 Max from Alibaba's Qwen team. It is a 2.4-trillion-parameter mixture-of-experts model that accepts text, image, and video input and returns text,...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen3 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen: Qwen3 235B A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.45,"output":3.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen/qwen3-vl-30b-a3b-thinking":{"id":"qwen/qwen3-vl-30b-a3b-thinking","name":"Qwen: Qwen3 VL 30B A3B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.29,"output":1}},"qwen/qwen-2.5-7b-instruct":{"id":"qwen/qwen-2.5-7b-instruct","name":"Qwen: Qwen2.5 7B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-10-16","last_updated":"2024-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":29491},"cost":{"input":0.1,"output":0.2}},"qwen/qwen3-vl-8b-instruct":{"id":"qwen/qwen3-vl-8b-instruct","name":"Qwen: Qwen3 VL 8B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.25,"output":0.75,"cache_read":0.12}},"qwen/qwen3-235b-a22b-2507":{"id":"qwen/qwen3-235b-a22b-2507","name":"Qwen: Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.1,"output":0.95,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.1625,"output":1.3},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.195,"output":1.56},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen-2.5-coder-32b-instruct":{"id":"qwen/qwen-2.5-coder-32b-instruct","name":"Qwen2.5 Coder 32B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.195,"output":0.975,"cache_read":0.039,"cache_write":0.24375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It is suited for coding, professional workflows, research, multimodal interaction, and long-running agent tasks, with flexible thinking that can be...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":235929},"cost":{"input":0.425,"output":2.55,"cache_read":0.085,"cache_write":0.53125},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Qwen3.7 Flash is a vision-language reasoning model from Alibaba. It is suited for multimodal agents, visual coding, search, and computer interaction, with strengths in object recognition, spatial understanding, and real-world...","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038},"canonical_model_id":"alibaba/qwen3.7-flash"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"Grok 4.7 is SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"Grok 4.6 is SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.20-multi-agent":{"id":"x-ai/grok-4.20-multi-agent","name":"SpaceXAI: Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Grok coding model for agentic engineering, edits, and codebase workflows","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":230400},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"SpaceXAI: Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":1800000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"Grok 4.5 is SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"bytedance-seed/seed-1.6":{"id":"bytedance-seed/seed-1.6","name":"ByteDance Seed: Seed 1.6 (retires Nov 11)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.25,"output":2},"canonical_model_id":"bytedance-seed/seed-1-6"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"Seed 2.0 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.25,"output":2},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-mini":{"id":"bytedance-seed/seed-2.0-mini","name":"Seed 2.0 Mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance-seed/seed-1.6-flash":{"id":"bytedance-seed/seed-1.6-flash","name":"ByteDance Seed: Seed 1.6 Flash (retires Nov 11)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.075,"output":0.3},"canonical_model_id":"bytedance-seed/seed-1-6-flash"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"Seed 2.0 Code","description":"Seed 2.0 Code is a model from ByteDance Seed optimized for agentic coding. It is suited for frontend development, multilingual programming tasks, and coding-agent workflows in tools such as Claude...","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":0.5,"output":3},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"bytedance-seed/seed-2-1-turbo":{"id":"bytedance-seed/seed-2-1-turbo","name":"ByteDance Seed: Seed 2.1 Turbo","description":"Seed 2.1 Turbo is a multimodal model from ByteDance Seed for coding and long-horizon agent workflows. It is suited for end-to-end software delivery, multi-step task execution, and understanding visual and...","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.5,"output":2.5},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":2.5,"output":5},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"scaledown/extract":{"id":"scaledown/extract","name":"ScaleDown: Extract","description":"Extract is a description-driven entity extraction model from ScaleDown. Instead of a fixed set of entity types, callers describe the fields they want in plain English through `provider.options.scaledown.entities`, and the...","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":900000},"cost":{"input":0.05,"output":0}},"scaledown/summarize":{"id":"scaledown/summarize","name":"ScaleDown: Summarize","description":"Summarize is an abstractive summarization model from ScaleDown. It condenses long text and documents into a fluent, faithful rewrite rather than lifting sentences from the source, and is suited for...","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":4000},"cost":{"input":0.05,"output":0}},"~z-ai/glm-flash-latest":{"id":"~z-ai/glm-flash-latest","name":"Z.ai: GLM Flash Latest","description":"This model always redirects to the latest model in the GLM Flash family.","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.0393,"output":0.1028,"cache_read":0.03}},"~z-ai/glm-latest":{"id":"~z-ai/glm-latest","name":"Z.ai: GLM Latest","description":"This model always redirects to the latest GLM model from Z.ai.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-19","last_updated":"2026-08-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.075,"output":4.295,"cache_read":0.075}},"bytedance/ui-tars-1.5-7b":{"id":"bytedance/ui-tars-1.5-7b","name":"ByteDance: UI-TARS 7B ","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["image","text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":2048},"cost":{"input":0.1,"output":0.2,"cache_read":0.1}},"writer/palmyra-x5":{"id":"writer/palmyra-x5","name":"Writer: Palmyra X5","description":"General-purpose chat model for instruction following, writing, and analysis","family":"palmyra","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1040000,"output":8192},"cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"cohere/command-a-plus":{"id":"cohere/command-a-plus","name":"Cohere: Command A+","description":"Command A+ is Cohere's flagship model for enterprise agentic workflows. It accepts text and image inputs with a 192K context window, supports native tool calling with strict tool schemas, structured...","family":"command-a","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04-01","release_date":"2026-05-20","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":192000,"output":64000},"cost":{"input":0.3,"output":1.5,"cache_read":0.15},"canonical_model_id":"cohere/command-a-plus-05-2026"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-r","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a"},"cohere/north-mini-code:free":{"id":"cohere/north-mini-code:free","name":"Cohere: North Mini Code (free)","description":"North Mini Code is Cohere's first agentic coding model and the debut of its North family. A sparse mixture-of-experts model with 30B total parameters and 3B active, it is optimized...","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"nex-agi/nex-n2.5-mini":{"id":"nex-agi/nex-n2.5-mini","name":"Nex AGI: Nex-N2.5-Mini","description":"Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. Its core strength is agentic coding within a visual feedback loop: it can explore codebases, implement multi-file...","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.025,"output":0.1,"cache_read":0.0025}},"nex-agi/nex-n2.5-pro":{"id":"nex-agi/nex-n2.5-pro","name":"Nex AGI: Nex-N2.5-Pro","description":"Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. Its core strength is agentic coding within a visual feedback loop: it can explore codebases, implement multi-file...","family":"agi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":235929},"cost":{"input":0.075,"output":0.25,"cache_read":0.015}},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"Morph: Morph V3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.9,"output":1.9},"canonical_model_id":"morph/morph-v3-large"},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"Morph: Morph V3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.8,"output":1.2},"canonical_model_id":"morph/morph-v3-fast"}}},"llmgateway-providers":{"id":"llmgateway-providers","env":["LLMGATEWAY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.llmgateway.io/v1","name":"LLM Gateway","doc":"https://llmgateway.io/docs","models":{"fireworks/deepseek-v4.1-flash":{"id":"fireworks/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Fireworks AI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"fireworks/kimi-k3":{"id":"fireworks/kimi-k3","name":"Kimi K3 (Fireworks AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1040384,"output":1040384},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"fireworks/kimi-k3-fast":{"id":"fireworks/kimi-k3-fast","name":"Kimi K3 Fast (Fireworks AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1040384,"output":1040384},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"fireworks/glm-5.3":{"id":"fireworks/glm-5.3","name":"GLM-5.3 (Fireworks AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"fireworks/glm-5.3-flash":{"id":"fireworks/glm-5.3-flash","name":"GLM-5.3 Flash (Fireworks AI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"vertex-openai/grok-4-7":{"id":"vertex-openai/grok-4-7","name":"Grok 4.7 (Vertex AI (OpenAI-compatible))","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":524288},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"vertex-openai/grok-4-6":{"id":"vertex-openai/grok-4-6","name":"Grok 4.6 (Vertex AI (OpenAI-compatible))","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"vertex-openai/grok-4-20-reasoning":{"id":"vertex-openai/grok-4-20-reasoning","name":"Grok 4.20 Reasoning (Vertex AI (OpenAI-compatible))","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"vertex-openai/grok-4-20-non-reasoning":{"id":"vertex-openai/grok-4-20-non-reasoning","name":"Grok 4.20 Non-Reasoning (Vertex AI (OpenAI-compatible))","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo V2.6 Pro (Xiaomi)","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo V2.6 Flash (Xiaomi)","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"luminal/deepseek-v4.1-flash":{"id":"luminal/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (luminal)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.135,"output":0.54,"cache_read":0.0027},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"ranoai/deepseek-v4-flash":{"id":"ranoai/deepseek-v4-flash","name":"DeepSeek V4 Flash (RanoAI)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"greenference/qwen3.6-27b":{"id":"greenference/qwen3.6-27b","name":"Qwen3.6 27B (greenference)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":8192},"cost":{"input":0.15,"output":1.35},"canonical_model_id":"alibaba/qwen3.6-27b"},"greenference/qwen3-14b":{"id":"greenference/qwen3-14b","name":"Qwen3 14B (greenference)","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192},"cost":{"input":0.076,"output":0.228}},"greenference/glm-5.3-flash":{"id":"greenference/glm-5.3-flash","name":"GLM-5.3 Flash (greenference)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.49,"cache_read":0.028},"canonical_model_id":"zhipuai/glm-5.3-flash"},"greenference/qwen3-30b-a3b":{"id":"greenference/qwen3-30b-a3b","name":"Qwen3 30B A3B (greenference)","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096},"cost":{"input":0.095,"output":0.285},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"aws-bedrock/llama-4-scout-17b-instruct":{"id":"aws-bedrock/llama-4-scout-17b-instruct","name":"Llama 4 Scout 17B Instruct (AWS Bedrock)","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"status":"deprecated","cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"aws-bedrock/claude-haiku-5-5":{"id":"aws-bedrock/claude-haiku-5-5","name":"Claude Haiku 5.5 (AWS Bedrock)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"aws-bedrock/claude-sonnet-4-5":{"id":"aws-bedrock/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (AWS Bedrock)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"aws-bedrock/claude-opus-4-5-20251101":{"id":"aws-bedrock/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (AWS Bedrock)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"aws-bedrock/grok-4-7":{"id":"aws-bedrock/grok-4-7","name":"Grok 4.7 (AWS Bedrock)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":524288},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"aws-bedrock/claude-sonnet-5":{"id":"aws-bedrock/claude-sonnet-5","name":"Claude Sonnet 5 (AWS Bedrock)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"aws-bedrock/grok-4-3":{"id":"aws-bedrock/grok-4-3","name":"Grok 4.3 (AWS Bedrock)","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"aws-bedrock/claude-opus-5":{"id":"aws-bedrock/claude-opus-5","name":"Claude Opus 5 (AWS Bedrock)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"aws-bedrock/claude-sonnet-4-6":{"id":"aws-bedrock/claude-sonnet-4-6","name":"Claude Sonnet 4.6 (AWS Bedrock)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"aws-bedrock/claude-sonnet-5-5":{"id":"aws-bedrock/claude-sonnet-5-5","name":"Claude Sonnet 5.5 (AWS Bedrock)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"aws-bedrock/claude-haiku-4-5":{"id":"aws-bedrock/claude-haiku-4-5","name":"Claude Haiku 4.5 (AWS Bedrock)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"aws-bedrock/claude-fable-5":{"id":"aws-bedrock/claude-fable-5","name":"Claude Fable 5 (AWS Bedrock)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"aws-bedrock/claude-fable-5-1":{"id":"aws-bedrock/claude-fable-5-1","name":"Claude Fable 5.1 (AWS Bedrock)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"aws-bedrock/claude-opus-4-6":{"id":"aws-bedrock/claude-opus-4-6","name":"Claude Opus 4.6 (AWS Bedrock)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"aws-bedrock/kimi-k3":{"id":"aws-bedrock/kimi-k3","name":"Kimi K3 (AWS Bedrock)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"moonshotai/kimi-k3"},"aws-bedrock/claude-sonnet-4-5-20250929":{"id":"aws-bedrock/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5 (2025-09-29) (AWS Bedrock)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"aws-bedrock/claude-opus-5-5":{"id":"aws-bedrock/claude-opus-5-5","name":"Claude Opus 5.5 (AWS Bedrock)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"aws-bedrock/claude-opus-4-8":{"id":"aws-bedrock/claude-opus-4-8","name":"Claude Opus 4.8 (AWS Bedrock)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"aws-bedrock/grok-4-6":{"id":"aws-bedrock/grok-4-6","name":"Grok 4.6 (AWS Bedrock)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"aws-bedrock/llama-3.1-70b-instruct":{"id":"aws-bedrock/llama-3.1-70b-instruct","name":"Llama 3.1 70B Instruct (AWS Bedrock)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"status":"deprecated","cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"aws-bedrock/llama-4-maverick-17b-instruct":{"id":"aws-bedrock/llama-4-maverick-17b-instruct","name":"Llama 4 Maverick 17B Instruct (AWS Bedrock)","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"status":"deprecated","cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"aws-bedrock/claude-haiku-4-5-20251001":{"id":"aws-bedrock/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5 (2025-10-01) (AWS Bedrock)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"aws-bedrock/claude-opus-4-7":{"id":"aws-bedrock/claude-opus-4-7","name":"Claude Opus 4.7 (AWS Bedrock)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"xai/grok-4-7":{"id":"xai/grok-4-7","name":"Grok 4.7 (xAI)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4-3":{"id":"xai/grok-4-3","name":"Grok 4.3 (xAI)","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-4-5":{"id":"xai/grok-4-5","name":"Grok 4.5 (xAI)","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"xai/grok-build-0-1":{"id":"xai/grok-build-0-1","name":"Grok Build 0.1 (xAI)","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4":{"id":"xai/grok-4","name":"Grok 4 (xAI)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":3,"output":15,"cache_read":0.75}},"xai/grok-4-20-beta-0309-reasoning":{"id":"xai/grok-4-20-beta-0309-reasoning","name":"Grok 4.20 Beta Reasoning (0309) (xAI)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4-6":{"id":"xai/grok-4-6","name":"Grok 4.6 (xAI)","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4-20-beta-0309-non-reasoning":{"id":"xai/grok-4-20-beta-0309-non-reasoning","name":"Grok 4.20 Beta Non-Reasoning (0309) (xAI)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"azure-anthropic/claude-sonnet-5":{"id":"azure-anthropic/claude-sonnet-5","name":"Claude Sonnet 5 (Azure Anthropic)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"azure-anthropic/claude-opus-5":{"id":"azure-anthropic/claude-opus-5","name":"Claude Opus 5 (Azure Anthropic)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"azure-anthropic/claude-sonnet-5-5":{"id":"azure-anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5 (Azure Anthropic)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"azure-anthropic/claude-fable-5":{"id":"azure-anthropic/claude-fable-5","name":"Claude Fable 5 (Azure Anthropic)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"azure-anthropic/claude-fable-5-1":{"id":"azure-anthropic/claude-fable-5-1","name":"Claude Fable 5.1 (Azure Anthropic)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"azure-anthropic/claude-opus-4-6":{"id":"azure-anthropic/claude-opus-4-6","name":"Claude Opus 4.6 (Azure Anthropic)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"azure-anthropic/claude-opus-5-5":{"id":"azure-anthropic/claude-opus-5-5","name":"Claude Opus 5.5 (Azure Anthropic)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"azure-anthropic/claude-opus-4-8":{"id":"azure-anthropic/claude-opus-4-8","name":"Claude Opus 4.8 (Azure Anthropic)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"azure-anthropic/claude-opus-4-7":{"id":"azure-anthropic/claude-opus-4-7","name":"Claude Opus 4.7 (Azure Anthropic)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"reka-ai/deepseek-v4.1-flash":{"id":"reka-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (reka-ai)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.1,"output":0.3,"cache_read":0.0025},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"reka-ai/deepseek-v4.1-flash-1017":{"id":"reka-ai/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (reka-ai)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"reka-ai/deepseek-v4-pro":{"id":"reka-ai/deepseek-v4-pro","name":"DeepSeek V4 Pro (reka-ai)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.9,"output":1.9,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-pro"},"reka-ai/glm-5.3":{"id":"reka-ai/glm-5.3","name":"GLM-5.3 (reka-ai)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.45,"output":1.8,"cache_read":0.135},"canonical_model_id":"zhipuai/glm-5.3"},"reka-ai/glm-5.3-flash":{"id":"reka-ai/glm-5.3-flash","name":"GLM-5.3 Flash (reka-ai)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-5.3-flash"},"reka-ai/reka-edge-2603":{"id":"reka-ai/reka-edge-2603","name":"Reka Edge 2603 (reka-ai)","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-09","last_updated":"2026-10-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":16384},"cost":{"input":0.1,"output":0.1}},"tencent/glm-5v-turbo":{"id":"tencent/glm-5v-turbo","name":"GLM-5V Turbo (Tencent Cloud)","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"tencent/kimi-k2.7-code":{"id":"tencent/kimi-k2.7-code","name":"Kimi K2.7 Code (Tencent Cloud)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"tencent/glm-5.2":{"id":"tencent/glm-5.2","name":"GLM-5.2 (Tencent Cloud)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"tencent/hy-mt2-plus":{"id":"tencent/hy-mt2-plus","name":"Hy-MT2 Plus (Tencent Cloud)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":4096},"cost":{"input":0.074,"output":0.295}},"tencent/deepseek-v4-pro":{"id":"tencent/deepseek-v4-pro","name":"DeepSeek V4 Pro (Tencent Cloud)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.435,"output":0.87,"cache_read":0.00363},"canonical_model_id":"deepseek/deepseek-v4-pro"},"tencent/glm-5.1":{"id":"tencent/glm-5.1","name":"GLM-5.1 (Tencent Cloud)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 Preview (Tencent Cloud)","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/minimax-m3":{"id":"tencent/minimax-m3","name":"MiniMax M3 (Tencent Cloud)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"tencent/kimi-k3":{"id":"tencent/kimi-k3","name":"Kimi K3 (Tencent Cloud)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"tencent/kimi-k2.6":{"id":"tencent/kimi-k2.6","name":"Kimi K2.6 (Tencent Cloud)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.858,"output":3.566,"cache_read":0.145},"canonical_model_id":"moonshotai/kimi-k2.6"},"tencent/minimax-m2.7":{"id":"tencent/minimax-m2.7","name":"MiniMax M2.7 (Tencent Cloud)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"tencent/glm-5":{"id":"tencent/glm-5","name":"GLM-5 (Tencent Cloud)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3 (Tencent Cloud)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"tencent/glm-5-turbo":{"id":"tencent/glm-5-turbo","name":"GLM-5 Turbo (Tencent Cloud)","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"tencent/deepseek-v4-flash":{"id":"tencent/deepseek-v4-flash","name":"DeepSeek V4 Flash (Tencent Cloud)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"tencent/kimi-k2.7-code-highspeed":{"id":"tencent/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed (Tencent Cloud)","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"novita/llama-4-scout-17b-instruct":{"id":"novita/llama-4-scout-17b-instruct","name":"Llama 4 Scout 17B Instruct (NovitaAI)","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.18,"output":0.59},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"novita/glm-4.5v":{"id":"novita/glm-4.5v","name":"GLM-4.5V (NovitaAI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"novita/qwen3-vl-235b-a22b-instruct":{"id":"novita/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct (NovitaAI)","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"novita/glm-4.6":{"id":"novita/glm-4.6","name":"GLM-4.6 (NovitaAI)","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.55,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"novita/kimi-k2.7-code":{"id":"novita/kimi-k2.7-code","name":"Kimi K2.7 Code (NovitaAI)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"novita/llama-3-70b-instruct":{"id":"novita/llama-3-70b-instruct","name":"Llama 3 70B Instruct (NovitaAI)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-04-18","last_updated":"2024-04-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8000},"cost":{"input":0.51,"output":0.74}},"novita/gemma-4-31b-it":{"id":"novita/gemma-4-31b-it","name":"Gemma 4 31B IT (NovitaAI)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"novita/llama-3.3-70b-instruct":{"id":"novita/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct (NovitaAI)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":120000},"cost":{"input":0.135,"output":0.4},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"novita/glm-5.2":{"id":"novita/glm-5.2","name":"GLM-5.2 (NovitaAI)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"novita/minimax-m3.1-flash":{"id":"novita/minimax-m3.1-flash","name":"MiniMax M3.1 Flash (NovitaAI)","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"novita/qwen3-vl-30b-a3b-instruct":{"id":"novita/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct (NovitaAI)","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.7},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"novita/qwen3-coder-30b-a3b-instruct":{"id":"novita/qwen3-coder-30b-a3b-instruct","name":"Qwen3 Coder 30B A3B Instruct (NovitaAI)","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":160000,"output":32768},"cost":{"input":0.07,"output":0.27},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"novita/deepseek-v4.1-flash":{"id":"novita/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (NovitaAI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"novita/step-3.7-flash":{"id":"novita/step-3.7-flash","name":"Step 3.7 Flash (NovitaAI)","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"novita/glm-4.7":{"id":"novita/glm-4.7","name":"GLM-4.7 (NovitaAI)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"novita/qwen3-235b-a22b-fp8":{"id":"novita/qwen3-235b-a22b-fp8","name":"Qwen3 235B A22B FP8 (NovitaAI)","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":20000},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"novita/deepseek-v4.1-flash-1017":{"id":"novita/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (NovitaAI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"novita/glm-5.1":{"id":"novita/glm-5.1","name":"GLM-5.1 (NovitaAI)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"novita/ling-3.0-flash":{"id":"novita/ling-3.0-flash","name":"InclusionAI Ling 3.0 Flash (NovitaAI)","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"novita/qwen3.8-2.4t-a95b":{"id":"novita/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B (NovitaAI)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"novita/minimax-m2.1":{"id":"novita/minimax-m2.1","name":"MiniMax M2.1 (NovitaAI)","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"novita/qwen3.8-flash":{"id":"novita/qwen3.8-flash","name":"Qwen3.8 Flash (NovitaAI)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"novita/qwen3-next-80b-a3b-instruct":{"id":"novita/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct (NovitaAI)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":1.5},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"novita/kimi-k3":{"id":"novita/kimi-k3","name":"Kimi K3 (NovitaAI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"novita/qwen3-235b-a22b-instruct-2507":{"id":"novita/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507 (NovitaAI)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.09,"output":0.58},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"novita/kimi-k2.6":{"id":"novita/kimi-k2.6","name":"Kimi K2.6 (NovitaAI)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.8,"output":3.4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"novita/minimax-m2.7":{"id":"novita/minimax-m2.7","name":"MiniMax M2.7 (NovitaAI)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"novita/qwen3.5-122b-a10b":{"id":"novita/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B (NovitaAI)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"novita/qwen3-coder-480b-a35b-instruct":{"id":"novita/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct (NovitaAI)","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"novita/qwen3-max":{"id":"novita/qwen3-max","name":"Qwen3 Max (NovitaAI)","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.845,"output":3.38},"canonical_model_id":"alibaba/qwen3-max"},"novita/glm-5":{"id":"novita/glm-5","name":"GLM-5 (NovitaAI)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"novita/glm-5.3":{"id":"novita/glm-5.3","name":"GLM-5.3 (NovitaAI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"novita/llama-4-maverick-17b-instruct":{"id":"novita/llama-4-maverick-17b-instruct","name":"Llama 4 Maverick 17B Instruct (NovitaAI)","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"novita/qwen3.7-max":{"id":"novita/qwen3.7-max","name":"Qwen3.7 Max (NovitaAI)","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":3.75,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.7-max"},"novita/glm-4.6v":{"id":"novita/glm-4.6v","name":"GLM-4.6V (NovitaAI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16000},"cost":{"input":0.3,"output":0.9,"cache_read":0.055},"canonical_model_id":"zhipuai/glm-4.6v"},"novita/hy3":{"id":"novita/hy3","name":"Hy3 (NovitaAI)","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"novita/qwen3-235b-a22b-thinking-2507":{"id":"novita/qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B A22B Thinking 2507 (NovitaAI)","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.3,"output":3},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"novita/minimax-m2.5":{"id":"novita/minimax-m2.5","name":"MiniMax M2.5 (NovitaAI)","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"novita/llama-3.2-3b-instruct":{"id":"novita/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct (NovitaAI)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-09-18","last_updated":"2024-09-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":32000},"cost":{"input":0.03,"output":0.05}},"novita/qwen3.6-35b-a3b":{"id":"novita/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B (NovitaAI)","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.248,"output":1.485},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"novita/deepseek-v3.2":{"id":"novita/deepseek-v3.2","name":"DeepSeek V3.2 (NovitaAI)","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.269,"output":0.4,"cache_read":0.1345},"canonical_model_id":"deepseek/deepseek-v3.2"},"novita/qwen3.8-max":{"id":"novita/qwen3.8-max","name":"Qwen3.8 Max (NovitaAI)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"novita/gemma-4-26b-a4b-it":{"id":"novita/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT (NovitaAI)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"novita/deepseek-v4-flash":{"id":"novita/deepseek-v4-flash","name":"DeepSeek V4 Flash (NovitaAI)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"novita/glm-5.3-flash":{"id":"novita/glm-5.3-flash","name":"GLM-5.3 Flash (NovitaAI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"novita/kimi-k2":{"id":"novita/kimi-k2","name":"Kimi K2 (NovitaAI)","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":0.57,"output":2.3}},"novita/qwen3-vl-235b-a22b-thinking":{"id":"novita/qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking (NovitaAI)","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.98,"output":3.95},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"novita/ernie-4.5-vl-424b-a47b":{"id":"novita/ernie-4.5-vl-424b-a47b","name":"ERNIE 4.5 VL 424B A47B (NovitaAI)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":123000,"output":16000},"cost":{"input":0.42,"output":1.25}},"novita/qwen35-397b-a17b":{"id":"novita/qwen35-397b-a17b","name":"Qwen3.5 397B A17B (NovitaAI)","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"novita/qwen3.8-27b":{"id":"novita/qwen3.8-27b","name":"Qwen3.8 27B (NovitaAI)","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.42,"output":3,"cache_read":0.085},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (DeepSeek)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":393216},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro (DeepSeek)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":393216},"cost":{"input":0.435,"output":0.87,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro"},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder 1.0 (Meta)","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1 (Meta)","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3 (Meta)","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2 (Meta)","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"runware/gemma-4-31b-it":{"id":"runware/gemma-4-31b-it","name":"Gemma 4 31B IT (Runware)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.102,"output":0.297,"cache_read":0.012},"canonical_model_id":"google/gemma-4-31b-it"},"runware/gpt-oss-120b":{"id":"runware/gpt-oss-120b","name":"GPT OSS 120B (Runware)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.032,"output":0.14,"cache_read":0.032},"canonical_model_id":"openai/gpt-oss-120b"},"runware/glm-5.2":{"id":"runware/glm-5.2","name":"GLM-5.2 (Runware)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":128000},"cost":{"input":0.8,"output":2.55,"cache_read":0.16},"canonical_model_id":"zhipuai/glm-5.2"},"runware/deepseek-v4.1-flash":{"id":"runware/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Runware)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"runware/deepseek-v4.1-flash-1017":{"id":"runware/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (runware)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.6,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"runware/deepseek-v4-pro":{"id":"runware/deepseek-v4-pro","name":"DeepSeek V4 Pro (Runware)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.961,"output":1.922,"cache_read":0.079},"canonical_model_id":"deepseek/deepseek-v4-pro"},"runware/kimi-k3":{"id":"runware/kimi-k3","name":"Kimi K3 (Runware)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"runware/kimi-k2.6":{"id":"runware/kimi-k2.6","name":"Kimi K2.6 (Runware)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.6,"output":3.05,"cache_read":0.13},"canonical_model_id":"moonshotai/kimi-k2.6"},"runware/glm-5.3":{"id":"runware/glm-5.3","name":"GLM-5.3 (Runware)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"runware/deepseek-v4-flash":{"id":"runware/deepseek-v4-flash","name":"DeepSeek V4 Flash (Runware)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.076,"output":0.153,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"runware/glm-5.3-flash":{"id":"runware/glm-5.3-flash","name":"GLM-5.3 Flash (Runware)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gonka24/minimax-m2.7":{"id":"gonka24/minimax-m2.7","name":"MiniMax M2.7 (Gonka24)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.08,"output":0.32,"cache_read":0.017},"canonical_model_id":"minimax/MiniMax-M2.7"},"google-ai-studio/gemini-4-argon":{"id":"google-ai-studio/gemini-4-argon","name":"Gemini 4 Argon (Google AI Studio)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"google-ai-studio/gemini-3-flash-preview":{"id":"google-ai-studio/gemini-3-flash-preview","name":"Gemini 3 Flash (Preview) (Google AI Studio)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google-ai-studio/gemini-3.8-flash":{"id":"google-ai-studio/gemini-3.8-flash","name":"Gemini 3.8 Flash (Google AI Studio)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.8-flash"},"google-ai-studio/gemini-3.1-pro-preview":{"id":"google-ai-studio/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview) (Google AI Studio)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google-ai-studio/gemini-pro-latest":{"id":"google-ai-studio/gemini-pro-latest","name":"Gemini Pro Latest (Google AI Studio)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-27","last_updated":"2026-02-27","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2}},"google-ai-studio/gemini-3.6-flash":{"id":"google-ai-studio/gemini-3.6-flash","name":"Gemini 3.6 Flash (Google AI Studio)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.6-flash"},"google-ai-studio/gemini-3.5-flash-lite":{"id":"google-ai-studio/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite (Google AI Studio)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google-ai-studio/gemini-3.1-flash-lite":{"id":"google-ai-studio/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite (Google AI Studio)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar (Perplexity)","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":130000,"output":4096},"cost":{"input":0.25,"output":2.5,"cache_write":0.25},"canonical_model_id":"perplexity/sonar"},"entrim/deepseek-v4.1-flash":{"id":"entrim/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (entrim)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"entrim/deepseek-v4.1-flash-1017":{"id":"entrim/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (entrim)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.8,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"entrim/glm-5.3":{"id":"entrim/glm-5.3","name":"GLM-5.3 (entrim)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.5,"output":2,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-5.3"},"entrim/glm-5.3-flash":{"id":"entrim/glm-5.3-flash","name":"GLM-5.3 Flash (entrim)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.0175},"canonical_model_id":"zhipuai/glm-5.3-flash"},"scx-ai/gpt-oss-120b":{"id":"scx-ai/gpt-oss-120b","name":"GPT OSS 120B (SCX.ai (Turbo))","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.17,"output":0.55},"canonical_model_id":"openai/gpt-oss-120b"},"scx-ai/minimax-m2.7":{"id":"scx-ai/minimax-m2.7","name":"MiniMax M2.7 (SCX.ai (Turbo))","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":196608},"cost":{"input":0.48,"output":1.79,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M2.7"},"scx-ai/qwen3-32b":{"id":"scx-ai/qwen3-32b","name":"Qwen3 32B (SCX.ai (Turbo))","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.36,"output":0.87},"canonical_model_id":"alibaba/qwen3-32b"},"baidu/glm-5.2":{"id":"baidu/glm-5.2","name":"GLM-5.2 (Baidu)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"baidu/deepseek-v4.1-flash":{"id":"baidu/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Baidu)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"baidu/deepseek-v4-pro":{"id":"baidu/deepseek-v4-pro","name":"DeepSeek V4 Pro (Baidu)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.32,"output":3.96,"cache_read":0.042},"canonical_model_id":"deepseek/deepseek-v4-pro"},"baidu/glm-5.1":{"id":"baidu/glm-5.1","name":"GLM-5.1 (Baidu)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"baidu/kimi-k2.6":{"id":"baidu/kimi-k2.6","name":"Kimi K2.6 (Baidu)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"baidu/glm-5.3":{"id":"baidu/glm-5.3","name":"GLM-5.3 (Baidu)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"baidu/deepseek-v4-flash":{"id":"baidu/deepseek-v4-flash","name":"DeepSeek V4 Flash (Baidu)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"alibaba/glm-5.2":{"id":"alibaba/glm-5.2","name":"GLM-5.2 (Alibaba Cloud)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.28},"canonical_model_id":"zhipuai/glm-5.2"},"alibaba/qwen3.7-plus":{"id":"alibaba/qwen3.7-plus","name":"Qwen3.7 Plus (Alibaba Cloud)","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"alibaba/qwen-plus":{"id":"alibaba/qwen-plus","name":"Qwen Plus (Alibaba Cloud)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32000},"cost":{"input":0.4,"output":1.2,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen-plus"},"alibaba/deepseek-v4.1-flash":{"id":"alibaba/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Alibaba Cloud)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"alibaba/deepseek-v4.1-flash-1017":{"id":"alibaba/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Alibaba Cloud)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"alibaba/qwen-max":{"id":"alibaba/qwen-max","name":"Qwen Max (Alibaba Cloud)","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"cost":{"input":1.6,"output":6.4},"canonical_model_id":"alibaba/qwen-max"},"alibaba/deepseek-v4-pro":{"id":"alibaba/deepseek-v4-pro","name":"DeepSeek V4 Pro (Alibaba Cloud)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":2.4,"output":4.8,"cache_read":0.2},"canonical_model_id":"deepseek/deepseek-v4-pro"},"alibaba/qwen3-vl-plus":{"id":"alibaba/qwen3-vl-plus","name":"Qwen3 VL Plus (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":1.6,"cache_read":0.04,"cache_write":0.25},"canonical_model_id":"alibaba/qwen3-vl-plus"},"alibaba/kimi-k2.5":{"id":"alibaba/kimi-k2.5","name":"Kimi K2.5 (Alibaba Cloud)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0.574,"output":3.011},"canonical_model_id":"moonshotai/kimi-k2.5"},"alibaba/qwen3.8-flash":{"id":"alibaba/qwen3.8-flash","name":"Qwen3.8 Flash (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"alibaba/qwen3.6-plus":{"id":"alibaba/qwen3.6-plus","name":"Qwen3.6 Plus (Alibaba Cloud)","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"alibaba/kimi-k3":{"id":"alibaba/kimi-k3","name":"Kimi K3 (Alibaba Cloud)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"alibaba/qwen3.5-122b-a10b":{"id":"alibaba/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"alibaba/qwen3.6-flash":{"id":"alibaba/qwen3.6-flash","name":"Qwen3.6 Flash (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.05,"cache_write":0.3125},"canonical_model_id":"alibaba/qwen3.6-flash"},"alibaba/glm-5":{"id":"alibaba/glm-5","name":"GLM-5 (Alibaba Cloud)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"cost":{"input":0.573,"output":2.58},"canonical_model_id":"zhipuai/glm-5"},"alibaba/glm-5.3":{"id":"alibaba/glm-5.3","name":"GLM-5.3 (Alibaba Cloud)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.28},"canonical_model_id":"zhipuai/glm-5.3"},"alibaba/qwen3.7-max":{"id":"alibaba/qwen3.7-max","name":"Qwen3.7 Max (Alibaba Cloud)","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":65536},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"alibaba/qwen-flash":{"id":"alibaba/qwen-flash","name":"Qwen Flash (Alibaba Cloud)","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.05,"output":0.4,"cache_read":0.01,"cache_write":0.0625},"canonical_model_id":"alibaba/qwen-flash"},"alibaba/qwen3.6-35b-a3b":{"id":"alibaba/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B (Alibaba Cloud)","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.375,"output":2.25},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen3.8 Max (Alibaba Cloud)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"alibaba/deepseek-v4-flash":{"id":"alibaba/deepseek-v4-flash","name":"DeepSeek V4 Flash (Alibaba Cloud)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.2,"output":0.4,"cache_read":0.04},"canonical_model_id":"deepseek/deepseek-v4-flash"},"alibaba/qwen3.5-35b-a3b":{"id":"alibaba/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"alibaba/qwen3.5-27b":{"id":"alibaba/qwen3.5-27b","name":"Qwen3.5 27B (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"},"alibaba/qwen-plus-latest":{"id":"alibaba/qwen-plus-latest","name":"Qwen Plus Latest (Alibaba Cloud)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-09-09","last_updated":"2024-09-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":0.4,"output":1.2,"cache_read":0.08,"cache_write":0.5}},"alibaba/qwen35-397b-a17b":{"id":"alibaba/qwen35-397b-a17b","name":"Qwen3.5 397B A17B (Alibaba Cloud)","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"alibaba/qwen3-coder-flash":{"id":"alibaba/qwen3-coder-flash","name":"Qwen3 Coder Flash (Alibaba Cloud)","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":997952,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"alibaba/qwen3-coder-flash"},"alibaba/qwen3.7-flash":{"id":"alibaba/qwen3.7-flash","name":"Qwen3.7 Flash (Alibaba Cloud)","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.0375},"canonical_model_id":"alibaba/qwen3.7-flash"},"meta-contributor/muse-spark-1.3-contributor":{"id":"meta-contributor/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor (Meta Contributor)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta-contributor/muse-spark-1.2-contributor":{"id":"meta-contributor/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor (Meta Contributor)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"canopywave/glm-5.2":{"id":"canopywave/glm-5.2","name":"GLM-5.2 (CanopyWave)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32768},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"canopywave/deepseek-v4-pro":{"id":"canopywave/deepseek-v4-pro","name":"DeepSeek V4 Pro (CanopyWave)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.74,"output":3.48,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4-pro"},"canopywave/kimi-k3":{"id":"canopywave/kimi-k3","name":"Kimi K3 (CanopyWave)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"canopywave/kimi-k2.6":{"id":"canopywave/kimi-k2.6","name":"Kimi K2.6 (CanopyWave)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"canopywave/deepseek-v4-flash":{"id":"canopywave/deepseek-v4-flash","name":"DeepSeek V4 Flash (CanopyWave)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash"},"cerebras/gemma-4-31b-it":{"id":"cerebras/gemma-4-31b-it","name":"Gemma 4 31B IT (Cerebras)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.99,"output":1.49},"canonical_model_id":"google/gemma-4-31b-it"},"cerebras/gpt-oss-120b":{"id":"cerebras/gpt-oss-120b","name":"GPT OSS 120B (Cerebras)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"cerebras/llama-3.3-70b-instruct":{"id":"cerebras/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct (Cerebras)","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.85,"output":1.2},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"cerebras/glm-4.7":{"id":"cerebras/glm-4.7","name":"GLM-4.7 (Cerebras)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":2.25,"output":2.75},"canonical_model_id":"zhipuai/glm-4.7"},"cerebras/qwen3-235b-a22b-instruct-2507":{"id":"cerebras/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507 (Cerebras)","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":8192},"cost":{"input":0.6,"output":1.2},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"google-vertex/gemini-4-argon":{"id":"google-vertex/gemini-4-argon","name":"Gemini 4 Argon (Google Vertex AI)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"google-vertex/gemini-3-flash-preview":{"id":"google-vertex/gemini-3-flash-preview","name":"Gemini 3 Flash (Preview) (Google Vertex AI)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google-vertex/gemini-3.5-flash":{"id":"google-vertex/gemini-3.5-flash","name":"Gemini 3.5 Flash (Google Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.08333,"input_audio":3},"canonical_model_id":"google/gemini-3.5-flash"},"google-vertex/gemini-3.8-flash":{"id":"google-vertex/gemini-3.8-flash","name":"Gemini 3.8 Flash (Google Vertex AI)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.8-flash"},"google-vertex/gemini-3.1-pro-preview":{"id":"google-vertex/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview) (Google Vertex AI)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google-vertex/gemini-3.7-flash":{"id":"google-vertex/gemini-3.7-flash","name":"Gemini 3.7 Flash (Google Vertex AI)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.7-flash"},"google-vertex/gemini-3.6-flash":{"id":"google-vertex/gemini-3.6-flash","name":"Gemini 3.6 Flash (Google Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.6-flash"},"google-vertex/gemini-3.5-flash-lite":{"id":"google-vertex/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite (Google Vertex AI)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google-vertex/gemini-3.1-flash-lite":{"id":"google-vertex/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite (Google Vertex AI)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"nebius/minimax-m3.1-flash":{"id":"nebius/minimax-m3.1-flash","name":"MiniMax M3.1 Flash (Nebius AI)","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"consensusprotocol/gemma-4-31b-it":{"id":"consensusprotocol/gemma-4-31b-it","name":"Gemma 4 31B IT (Consensus Protocol)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.25,"cache_read":0.01},"canonical_model_id":"google/gemma-4-31b-it"},"consensusprotocol/deepseek-v4.1-flash":{"id":"consensusprotocol/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Consensus Protocol)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.15,"output":0.55,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"consensusprotocol/gpt-oss-20b":{"id":"consensusprotocol/gpt-oss-20b","name":"GPT OSS 20B (Consensus Protocol)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.04,"output":0.19,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-20b"},"consensusprotocol/minimax-m3":{"id":"consensusprotocol/minimax-m3","name":"MiniMax M3 (Consensus Protocol)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"consensusprotocol/Qwen3.8-27B":{"id":"consensusprotocol/Qwen3.8-27B","name":"Qwen3.8 27B (Consensus Protocol)","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":32768},"cost":{"input":0.08,"output":0.35,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"consensusprotocol/mimo-v2.6-flash":{"id":"consensusprotocol/mimo-v2.6-flash","name":"MiMo V2.6 Flash (Consensus Protocol)","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.12,"output":0.25,"cache_read":0.0025},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"consensusprotocol/deepseek-v4-flash":{"id":"consensusprotocol/deepseek-v4-flash","name":"DeepSeek V4 Flash (Consensus Protocol)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.1,"output":0.2,"cache_read":0.005},"canonical_model_id":"deepseek/deepseek-v4-flash"},"consensusprotocol/glm-5.3-flash":{"id":"consensusprotocol/glm-5.3-flash","name":"GLM-5.3 Flash (Consensus Protocol)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.25,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-5.3-flash"},"runpod/kimi-k3":{"id":"runpod/kimi-k3","name":"Kimi K3 (Runpod)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"groq/gpt-oss-120b":{"id":"groq/gpt-oss-120b","name":"GPT OSS 120B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.15,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"groq/gpt-oss-20b":{"id":"groq/gpt-oss-20b","name":"GPT OSS 20B (Groq)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32766},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-oss-20b"},"scx-ai-gp/kimi-k2.7-code":{"id":"scx-ai-gp/kimi-k2.7-code","name":"Kimi K2.7 Code (SCX.ai)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.737,"output":3.08,"cache_read":0.143},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"scx-ai-gp/glm-5.2":{"id":"scx-ai-gp/glm-5.2","name":"GLM-5.2 (SCX.ai)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.605,"output":2.123,"cache_read":0.154},"canonical_model_id":"zhipuai/glm-5.2"},"scx-ai-gp/deepseek-v4.1-flash":{"id":"scx-ai-gp/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (SCX.ai)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.01},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"scx-ai-gp/kimi-k3":{"id":"scx-ai-gp/kimi-k3","name":"Kimi K3 (SCX.ai)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3.135,"output":15.675,"cache_read":0.319},"canonical_model_id":"moonshotai/kimi-k3"},"scx-ai-gp/glm-5.2-fast":{"id":"scx-ai-gp/glm-5.2-fast","name":"GLM-5.2 Turbo (SCX.ai)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":2.156,"output":6.776,"cache_read":0.429},"canonical_model_id":"zhipuai/glm-5.2"},"scx-ai-gp/glm-5.3":{"id":"scx-ai-gp/glm-5.3","name":"GLM-5.3 (SCX.ai)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.924,"output":2.904,"cache_read":0.1716},"canonical_model_id":"zhipuai/glm-5.3"},"scx-ai-gp/glm-5.3-fast":{"id":"scx-ai-gp/glm-5.3-fast","name":"GLM 5.3 Fast (SCX.ai)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":2.8,"output":8.8,"cache_read":0.56},"canonical_model_id":"zhipuai/glm-5.3"},"scx-ai-gp/qwen3.8-max":{"id":"scx-ai-gp/qwen3.8-max","name":"Qwen3.8 Max (SCX.ai)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.54,"output":4.631,"cache_read":0.198,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"scx-ai-gp/glm-5.3-flash":{"id":"scx-ai-gp/glm-5.3-flash","name":"GLM-5.3 Flash (SCX.ai)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.088,"output":0.25,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-5.3-flash"},"embercloud/glm-4.7-flash":{"id":"embercloud/glm-4.7-flash","name":"GLM-4.7 Flash (EmberCloud)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131000},"cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"embercloud/glm-4.5-air":{"id":"embercloud/glm-4.5-air","name":"GLM-4.5 Air (EmberCloud)","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":96000},"cost":{"input":0.13,"output":0.85,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-4.5-air"},"embercloud/glm-5.2":{"id":"embercloud/glm-5.2","name":"GLM-5.2 (EmberCloud)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131000},"cost":{"input":1.26,"output":3.96,"cache_read":0.234},"canonical_model_id":"zhipuai/glm-5.2"},"embercloud/glm-4.7":{"id":"embercloud/glm-4.7","name":"GLM-4.7 (EmberCloud)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131000},"cost":{"input":0.38,"output":1.98,"cache_read":0.19},"canonical_model_id":"zhipuai/glm-4.7"},"embercloud/glm-5.1":{"id":"embercloud/glm-5.1","name":"GLM-5.1 (EmberCloud)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131000},"cost":{"input":0.931,"output":2.93,"cache_read":0.173},"canonical_model_id":"zhipuai/glm-5.1"},"embercloud/kimi-k2.5":{"id":"embercloud/kimi-k2.5","name":"Kimi K2.5 (EmberCloud)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.405,"output":1.98,"cache_read":0.225},"canonical_model_id":"moonshotai/kimi-k2.5"},"embercloud/qwen3-coder-next":{"id":"embercloud/qwen3-coder-next","name":"Qwen3 Coder Next (EmberCloud)","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.108,"output":0.675,"cache_read":0.06},"canonical_model_id":"alibaba/qwen3-coder-next"},"embercloud/glm-4.5":{"id":"embercloud/glm-4.5","name":"GLM-4.5 (EmberCloud)","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":96000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"embercloud/glm-5":{"id":"embercloud/glm-5","name":"GLM-5 (EmberCloud)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":131000},"cost":{"input":0.72,"output":2.3,"cache_read":0.144},"canonical_model_id":"zhipuai/glm-5"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini (OpenAI)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1 (OpenAI)","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4 (OpenAI)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra (OpenAI)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1 (OpenAI)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini (OpenAI)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro (OpenAI)","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5 (OpenAI)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna (OpenAI)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano (OpenAI)","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-live-transcribe":{"id":"openai/gpt-live-transcribe","type":"transcription","name":"GPT Live Transcribe (OpenAI)","description":"Low-latency streaming speech-to-text model with keyword and language hints","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-live-transcribe"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol (OpenAI)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini (OpenAI)","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro (OpenAI)","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex (OpenAI)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2 (OpenAI)","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra (OpenAI)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol (OpenAI)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro (OpenAI)","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro (OpenAI)","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-4o-transcribe":{"id":"openai/gpt-4o-transcribe","name":"GPT-4o Transcribe (OpenAI)","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"cost":{"input":2.5,"output":10,"input_audio":6}},"openai/gpt-transcribe":{"id":"openai/gpt-transcribe","type":"transcription","name":"GPT Transcribe (OpenAI)","description":"High-accuracy speech-to-text model for audio files and committed Realtime turns","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-transcribe"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano (OpenAI)","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o (OpenAI)","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o3":{"id":"openai/o3","name":"o3 (OpenAI)","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4o-mini-transcribe":{"id":"openai/gpt-4o-mini-transcribe","name":"GPT-4o Mini Transcribe (OpenAI)","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"cost":{"input":1.25,"output":5,"input_audio":3}},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna (OpenAI)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o Mini (OpenAI)","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5 (OpenAI)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol (OpenAI)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"GLM-4.5V (Z AI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6 (Z AI)","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5 Air (Z AI)","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2 (Z AI)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7 (Z AI)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1 (Z AI)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4-32b-0414-128k":{"id":"zai/glm-4-32b-0414-128k","name":"GLM-4 32B (0414-128k) (Z AI)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":0.1,"output":0.1}},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5 (Z AI)","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-4.5-airx":{"id":"zai/glm-4.5-airx","name":"GLM-4.5 AirX (Z AI)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":1.1,"output":4.5,"cache_read":0.22}},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5 (Z AI)","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131100},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM-4.7 FlashX (Z AI)","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3 (Z AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-4.6v":{"id":"zai/glm-4.6v","name":"GLM-4.6V (Z AI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.3,"output":0.9,"cache_read":0.05},"canonical_model_id":"zhipuai/glm-4.6v"},"zai/glm-4.5-x":{"id":"zai/glm-4.5-x","name":"GLM-4.5 X (Z AI)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":2.2,"output":8.9,"cache_read":0.45}},"zai/glm-4.6v-flashx":{"id":"zai/glm-4.6v-flashx","name":"GLM-4.6V FlashX (Z AI)","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.04,"output":0.4,"cache_read":0.004}},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3 Flash (Z AI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"inference.net/minimax-m3.1-flash":{"id":"inference.net/minimax-m3.1-flash","name":"MiniMax M3.1 Flash (Inference.net)","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.1,"output":0.5,"cache_read":0.005},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"inference.net/deepseek-v4.1-flash-1017":{"id":"inference.net/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Inference.net)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.3,"output":1,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"inference.net/glm-5.3":{"id":"inference.net/glm-5.3","name":"GLM-5.3 (Inference.net)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.9,"output":3,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-5.3"},"inference.net/glm-5.3-flash":{"id":"inference.net/glm-5.3-flash","name":"GLM-5.3 Flash (Inference.net)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":0.09,"output":0.28,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-5.3-flash"},"inference.net/llama-3.2-11b-instruct":{"id":"inference.net/llama-3.2-11b-instruct","name":"Llama 3.2 11B Instruct (Inference.net)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":0.07,"output":0.33}},"azure/gpt-4.1-mini":{"id":"azure/gpt-4.1-mini","name":"GPT-4.1 Mini (Azure)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"azure/gpt-oss-120b":{"id":"azure/gpt-oss-120b","name":"GPT OSS 120B (Azure)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"azure/gpt-5.1":{"id":"azure/gpt-5.1","name":"GPT-5.1 (Azure)","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"azure/gpt-5.4":{"id":"azure/gpt-5.4","name":"GPT-5.4 (Azure)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"azure/gpt-6-astra":{"id":"azure/gpt-6-astra","name":"GPT-6 Astra (Azure)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"azure/gpt-4.1":{"id":"azure/gpt-4.1","name":"GPT-4.1 (Azure)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"azure/gpt-5.1-codex":{"id":"azure/gpt-5.1-codex","name":"GPT-5.1 Codex (Azure)","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-codex"},"azure/gpt-5-mini":{"id":"azure/gpt-5-mini","name":"GPT-5 Mini (Azure)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"azure/gpt-5.5":{"id":"azure/gpt-5.5","name":"GPT-5.5 (Azure)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"azure/gpt-6-luna":{"id":"azure/gpt-6-luna","name":"GPT-6 Luna (Azure)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"azure/gpt-5-nano":{"id":"azure/gpt-5-nano","name":"GPT-5 Nano (Azure)","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"azure/gpt-6-sol":{"id":"azure/gpt-6-sol","name":"GPT-6 Sol (Azure)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"azure/gpt-5.4-mini":{"id":"azure/gpt-5.4-mini","name":"GPT-5.4 Mini (Azure)","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"azure/gpt-5.3-codex":{"id":"azure/gpt-5.3-codex","name":"GPT-5.3 Codex (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"azure/gpt-5.2":{"id":"azure/gpt-5.2","name":"GPT-5.2 (Azure)","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"azure/gpt-5.6-terra":{"id":"azure/gpt-5.6-terra","name":"GPT-5.6 Terra (Azure)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"azure/gpt-6.1-sol":{"id":"azure/gpt-6.1-sol","name":"GPT-6.1 Sol (Azure)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"azure/gpt-5.4-pro":{"id":"azure/gpt-5.4-pro","name":"GPT-5.4 Pro (Azure)","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.4-pro"},"azure/gpt-5.4-nano":{"id":"azure/gpt-5.4-nano","name":"GPT-5.4 Nano (Azure)","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"azure/gpt-5.1-codex-mini":{"id":"azure/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini (Azure)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"azure/gpt-4o":{"id":"azure/gpt-4o","name":"GPT-4o (Azure)","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"azure/gpt-5.6-luna":{"id":"azure/gpt-5.6-luna","name":"GPT-5.6 Luna (Azure)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"azure/gpt-5.2-codex":{"id":"azure/gpt-5.2-codex","name":"GPT-5.2 Codex (Azure)","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"azure/gpt-5":{"id":"azure/gpt-5","name":"GPT-5 (Azure)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"azure/gpt-5.6-sol":{"id":"azure/gpt-5.6-sol","name":"GPT-5.6 Sol (Azure)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5 (Anthropic)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (Anthropic)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (Anthropic)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5 (Anthropic)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5 (Anthropic)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6 (Anthropic)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5 (Anthropic)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (Anthropic)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5 (Anthropic)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1 (Anthropic)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6 (Anthropic)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-4-5-20250929":{"id":"anthropic/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5 (2025-09-29) (Anthropic)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5 (Anthropic)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8 (Anthropic)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-haiku-4-5-20251001":{"id":"anthropic/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5 (2025-10-01) (Anthropic)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7 (Anthropic)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max (Sakana AI)","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/fugu-ultra-v2.0":{"id":"sakana/fugu-ultra-v2.0","name":"Fugu Ultra v2.0 (Sakana AI)","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra (Sakana AI)","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"azure-ai-foundry/grok-4-1-fast-reasoning":{"id":"azure-ai-foundry/grok-4-1-fast-reasoning","name":"Grok 4.1 Fast Reasoning (Azure AI Foundry)","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"azure-ai-foundry/grok-4-3":{"id":"azure-ai-foundry/grok-4-3","name":"Grok 4.3 (Azure AI Foundry)","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":20000,"output":8192},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"azure-ai-foundry/grok-4-1-fast-non-reasoning":{"id":"azure-ai-foundry/grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast Non-Reasoning (Azure AI Foundry)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5}},"together-ai/gpt-oss-120b":{"id":"together-ai/gpt-oss-120b","name":"GPT OSS 120B (Together AI)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"together-ai/minimax-m3.1-flash":{"id":"together-ai/minimax-m3.1-flash","name":"MiniMax M3.1 Flash (Together AI)","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"together-ai/deepseek-v4.1-flash":{"id":"together-ai/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (Together AI)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"together-ai/glm-4.7":{"id":"together-ai/glm-4.7","name":"GLM-4.7 (Together AI)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0.45,"output":2},"canonical_model_id":"zhipuai/glm-4.7"},"together-ai/muse-glimmer-30b":{"id":"together-ai/muse-glimmer-30b","name":"Muse Glimmer 30B (Together AI)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"together-ai/deepseek-v4.1-flash-1017":{"id":"together-ai/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (Together AI)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"together-ai/deepseek-v4-pro":{"id":"together-ai/deepseek-v4-pro","name":"DeepSeek V4 Pro (Together AI)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":163840},"cost":{"input":1.32,"output":3.96,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro"},"together-ai/minimax-m3":{"id":"together-ai/minimax-m3","name":"MiniMax M3 (Together AI)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"together-ai/qwen3.8-2.4t-a95b":{"id":"together-ai/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B (Together AI)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1010000,"output":909000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"together-ai/kimi-k3":{"id":"together-ai/kimi-k3","name":"Kimi K3 (Together AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"together-ai/glm-5.3":{"id":"together-ai/glm-5.3","name":"GLM-5.3 (Together AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"together-ai/glm-5.3-flash":{"id":"together-ai/glm-5.3-flash","name":"GLM-5.3 Flash (Together AI)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":943717},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 Highspeed (MiniMax)","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.6,"output":2.4,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax M3.1 Flash (MiniMax)","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-text-01":{"id":"minimax/minimax-text-01","name":"MiniMax Text 01 (MiniMax)","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.2,"output":1.1}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3 (MiniMax)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1 (MiniMax)","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.27,"output":1.1},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7 (MiniMax)","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1-lightning":{"id":"minimax/minimax-m2.1-lightning","name":"MiniMax M2.1 Lightning (MiniMax)","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":131072},"cost":{"input":0.12,"output":0.48}},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax M2.5 Highspeed (MiniMax)","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.6,"output":2.4,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5 (MiniMax)","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2 (MiniMax)","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.2,"output":1,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2"},"mistral/mistral-small-2506":{"id":"mistral/mistral-small-2506","name":"Mistral Small 3.2 (Mistral AI)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"mistral/mistral-small-2506"},"mistral/ministral-14b-2512":{"id":"mistral/ministral-14b-2512","name":"Ministral 14B (Mistral AI)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"mistral/ministral-14b-2512"},"mistral/devstral-2512":{"id":"mistral/devstral-2512","name":"Devstral 2 (Mistral AI)","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-2512"},"mistral/codestral-2508":{"id":"mistral/codestral-2508","name":"Codestral (Mistral AI)","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-2508"},"mistral/ministral-8b-2512":{"id":"mistral/ministral-8b-2512","name":"Ministral 8B (Mistral AI)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.15},"canonical_model_id":"mistral/ministral-8b-2512"},"mistral/mistral-large-4":{"id":"mistral/mistral-large-4","name":"Mistral Large 4 (Mistral AI)","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistral/glm-5.3":{"id":"mistral/glm-5.3","name":"GLM-5.3 (Mistral AI)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large Latest (Mistral AI)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"cost":{"input":4,"output":12},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/mistral-small-2603":{"id":"mistral/mistral-small-2603","name":"Mistral Small 4 (Mistral AI)","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-2603"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3 (Mistral AI)","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/mistral-medium-3-5":{"id":"mistral/mistral-medium-3-5","name":"Mistral Medium 3.5 (Mistral AI)","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistral/ministral-3b-2512":{"id":"mistral/ministral-3b-2512","name":"Ministral 3B (Mistral AI)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"mistral/ministral-3b-2512"},"deepinfra/mimo-v2.6-pro":{"id":"deepinfra/mimo-v2.6-pro","name":"MiMo V2.6 Pro (DeepInfra)","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"deepinfra/qwen3.5-9b":{"id":"deepinfra/qwen3.5-9b","name":"Qwen3.5 9B (DeepInfra)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"deepinfra/qwen3-vl-235b-a22b-instruct":{"id":"deepinfra/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct (DeepInfra)","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"deepinfra/gemma-4-31b-it":{"id":"deepinfra/gemma-4-31b-it","name":"Gemma 4 31B IT (DeepInfra)","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"deepinfra/qwen3-vl-30b-a3b-instruct":{"id":"deepinfra/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct (DeepInfra)","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"deepinfra/deepseek-v4.1-flash":{"id":"deepinfra/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash (DeepInfra)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepinfra/ling-3.0-flash-vl":{"id":"deepinfra/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL (DeepInfra)","description":"Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"deepinfra/muse-glimmer-30b":{"id":"deepinfra/muse-glimmer-30b","name":"Muse Glimmer 30B (DeepInfra)","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.3,"output":1.2,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"deepinfra/deepseek-v4.1-flash-1017":{"id":"deepinfra/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017 (DeepInfra)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepinfra/deepseek-v4-pro":{"id":"deepinfra/deepseek-v4-pro","name":"DeepSeek V4 Pro (DeepInfra)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepinfra/ling-3.0-flash":{"id":"deepinfra/ling-3.0-flash","name":"InclusionAI Ling 3.0 Flash (DeepInfra)","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash"},"deepinfra/nemotron-3-ultra-550b":{"id":"deepinfra/nemotron-3-ultra-550b","name":"Nemotron 3 Ultra 550B (DeepInfra)","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.2,"cache_read":0.1}},"deepinfra/mimo-v2.6-flash":{"id":"deepinfra/mimo-v2.6-flash","name":"MiMo V2.6 Flash (DeepInfra)","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"deepinfra/granite-4.2-8b":{"id":"deepinfra/granite-4.2-8b","name":"Granite 4.2 8B (DeepInfra)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964},"cost":{"input":0.06,"output":0.25,"cache_read":0.015},"canonical_model_id":"ibm/granite-4.2-8b"},"deepinfra/deepseek-v3.2":{"id":"deepinfra/deepseek-v3.2","name":"DeepSeek V3.2 (DeepInfra)","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":160000,"output":65536},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepinfra/gemma-4-26b-a4b-it":{"id":"deepinfra/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT (DeepInfra)","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"deepinfra/deepseek-v4-flash":{"id":"deepinfra/deepseek-v4-flash","name":"DeepSeek V4 Flash (DeepInfra)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.08,"output":0.18,"cache_read":0.016},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code (Moonshot AI)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3.1-pro":{"id":"moonshot/kimi-k3.1-pro","name":"Kimi K3.1 Pro (Moonshot AI)","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshot/kimi-k2.5":{"id":"moonshot/kimi-k2.5","name":"Kimi K2.5 (Moonshot AI)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3 (Moonshot AI)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6 (Moonshot AI)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k3.1-pro-highspeed":{"id":"moonshot/kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed (Moonshot AI)","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4}},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed (Moonshot AI)","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"vertex-anthropic/claude-sonnet-4-5":{"id":"vertex-anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (Vertex AI (Anthropic))","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"vertex-anthropic/claude-opus-4-5-20251101":{"id":"vertex-anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (Vertex AI (Anthropic))","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"vertex-anthropic/claude-sonnet-5":{"id":"vertex-anthropic/claude-sonnet-5","name":"Claude Sonnet 5 (Vertex AI (Anthropic))","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"vertex-anthropic/claude-sonnet-4-6":{"id":"vertex-anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6 (Vertex AI (Anthropic))","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"vertex-anthropic/claude-haiku-4-5":{"id":"vertex-anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (Vertex AI (Anthropic))","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"vertex-anthropic/claude-opus-4-6":{"id":"vertex-anthropic/claude-opus-4-6","name":"Claude Opus 4.6 (Vertex AI (Anthropic))","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"vertex-anthropic/claude-opus-4-7":{"id":"vertex-anthropic/claude-opus-4-7","name":"Claude Opus 4.7 (Vertex AI (Anthropic))","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"aws-mantle/gpt-6-astra":{"id":"aws-mantle/gpt-6-astra","name":"GPT-6 Astra (AWS Mantle)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"aws-mantle/gpt-5.6-terra":{"id":"aws-mantle/gpt-5.6-terra","name":"GPT-5.6 Terra (AWS Mantle)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":921600,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"openai/gpt-5.6-terra"},"aws-mantle/gpt-5.6-luna":{"id":"aws-mantle/gpt-5.6-luna","name":"GPT-5.6 Luna (AWS Mantle)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":921600,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275},"canonical_model_id":"openai/gpt-5.6-luna"},"aws-mantle/gpt-5.6-sol":{"id":"aws-mantle/gpt-5.6-sol","name":"GPT-5.6 Sol (AWS Mantle)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":921600,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5},"canonical_model_id":"openai/gpt-5.6-sol"},"bytedance/gpt-oss-120b":{"id":"bytedance/gpt-oss-120b","name":"GPT OSS 120B (ByteDance)","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.1,"output":0.5,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-120b"},"bytedance/glm-5.2":{"id":"bytedance/glm-5.2","name":"GLM-5.2 (ByteDance)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"bytedance/glm-4.7":{"id":"bytedance/glm-4.7","name":"GLM-4.7 (ByteDance)","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"bytedance/deepseek-v4-pro":{"id":"bytedance/deepseek-v4-pro","name":"DeepSeek V4 Pro (ByteDance)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"bytedance/seed-1-8-251228":{"id":"bytedance/seed-1-8-251228","name":"Seed 1.8 (251228) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/seed-1-6-250615":{"id":"bytedance/seed-1-6-250615","name":"Seed 1.6 (250615) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-06-25","last_updated":"2025-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/seed-1-6-flash-250715":{"id":"bytedance/seed-1-6-flash-250715","name":"Seed 1.6 Flash (250715) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.07,"output":0.3,"cache_read":0.015}},"bytedance/seed-2-0-mini-260428":{"id":"bytedance/seed-2-0-mini-260428","name":"Seed 2.0 Mini (260428) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.1,"output":0.4,"cache_read":0.02}},"bytedance/seed-2-0-code-preview-260328":{"id":"bytedance/seed-2-0-code-preview-260328","name":"Seed 2.0 Code Preview (260328) (ByteDance)","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"bytedance/seed-2-0-lite-260428":{"id":"bytedance/seed-2-0-lite-260428","name":"Seed 2.0 Lite (260428) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/seed-1-6-250915":{"id":"bytedance/seed-1-6-250915","name":"Seed 1.6 (250915) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.25,"output":2,"cache_read":0.05}},"bytedance/deepseek-v3.2":{"id":"bytedance/deepseek-v3.2","name":"DeepSeek V3.2 (ByteDance)","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.28,"output":0.42,"cache_read":0.056},"canonical_model_id":"deepseek/deepseek-v3.2"},"bytedance/deepseek-v4-flash":{"id":"bytedance/deepseek-v4-flash","name":"DeepSeek V4 Flash (ByteDance)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash"},"bytedance/seed-2-0-pro-260328":{"id":"bytedance/seed-2-0-pro-260328","name":"Seed 2.0 Pro (260328) (ByteDance)","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1}},"quartz/gemini-3.1-pro-preview":{"id":"quartz/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview) (Quartz)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"}}},"lilac":{"id":"lilac","env":["LILAC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.getlilac.com/v1","name":"Lilac","doc":"https://docs.getlilac.com/inference/models","models":{"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":262100},"status":"deprecated","cost":{"input":0.11,"output":0.35},"canonical_model_id":"google/gemma-4-31b-it"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.7,"output":3.5,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimaxai/minimax-m3":{"id":"minimaxai/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"status":"deprecated","cost":{"input":0.28,"output":1.1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"zai-org/glm-5.2":{"id":"zai-org/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":524288},"status":"deprecated","cost":{"input":0.9,"output":3,"cache_read":0.27},"canonical_model_id":"zhipuai/glm-5.2"}}},"groq":{"id":"groq","env":["GROQ_API_KEY"],"npm":"@ai-sdk/groq","name":"Groq","doc":"https://console.groq.com/docs/models","models":{"allam-2-7b":{"id":"allam-2-7b","name":"ALLaM-2-7b","description":"ALLaM-2-7b instruction tuned model by SDAIA","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-01-23","last_updated":"2025-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"sdaia/allam-2-7b"},"llama-3.3-70b-versatile":{"id":"llama-3.3-70b-versatile","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"whisper-large-v3-turbo":{"id":"whisper-large-v3-turbo","type":"transcription","name":"Whisper Large V3 Turbo","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"canonical_model_id":"openai/whisper-large-v3-turbo"},"llama-3.1-8b-instant":{"id":"llama-3.1-8b-instant","name":"Llama 3.1 8B","description":"Compact Llama instruction model for fast chat and local deployment","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.05,"output":0.08},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"whisper-large-v3":{"id":"whisper-large-v3","type":"transcription","name":"Whisper","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"canonical_model_id":"openai/whisper-large-v3"},"meta-llama/llama-prompt-guard-2-86m":{"id":"meta-llama/llama-prompt-guard-2-86m","name":"Prompt Guard 2 86M","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"status":"beta","cost":{"input":0.04,"output":0.04},"canonical_model_id":"meta/llama-prompt-guard-2-86m"},"meta-llama/llama-prompt-guard-2-22m":{"id":"meta-llama/llama-prompt-guard-2-22m","name":"Llama Prompt Guard 2 22M","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"status":"beta","cost":{"input":0.03,"output":0.03},"canonical_model_id":"meta/llama-prompt-guard-2-22m"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"Safety GPT OSS 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"beta","cost":{"input":0.075,"output":0.3,"cache_read":0.0375},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"canopylabs/orpheus-arabic-saudi":{"id":"canopylabs/orpheus-arabic-saudi","type":"speech","name":"Canopy Labs Orpheus Arabic Saudi","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"canopylabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":4000,"output":50000},"status":"beta","provider":{"npm":"@ai-sdk/openai","api":"https://api.groq.com/openai/v1"},"canonical_model_id":"canopylabs/orpheus-arabic-saudi"},"canopylabs/orpheus-v1-english":{"id":"canopylabs/orpheus-v1-english","type":"speech","name":"Canopy Labs Orpheus V1 English","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"canopylabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-19","last_updated":"2025-12-19","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":4000,"output":50000},"status":"beta","provider":{"npm":"@ai-sdk/openai","api":"https://api.groq.com/openai/v1"},"canonical_model_id":"canopylabs/orpheus-v1-english"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","default"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":0.6,"output":3,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","default","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"status":"beta","cost":{"input":0.8,"output":4},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"azure-cognitive-services":{"id":"azure-cognitive-services","env":["AZURE_COGNITIVE_SERVICES_RESOURCE_NAME","AZURE_COGNITIVE_SERVICES_API_KEY"],"npm":"@ai-sdk/azure","name":"Azure Cognitive Services","doc":"https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/models","models":{"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"GPT-Image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":32,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1.5"},"o1":{"id":"o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5}},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.71,"output":0.71}},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}},"gpt-image-2.5-flare":{"id":"gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"phi-4-reasoning-plus":{"id":"phi-4-reasoning-plus","name":"Phi-4-reasoning-plus","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek-V4.1-Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"llama-4-scout-17b-16e-instruct":{"id":"llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B 16E Instruct","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.78}},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek-V4-Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":1.74,"output":3.48},"canonical_model_id":"deepseek/deepseek-v4-pro"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"status":"beta","provider":{"shape":"completions"},"canonical_model_id":"xai/grok-4.7"},"cohere-embed-v3-multilingual":{"id":"cohere-embed-v3-multilingual","type":"embedding","name":"Embed v3 Multilingual","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"deepseek-v3.2-speciale":{"id":"deepseek-v3.2-speciale","name":"DeepSeek-V3.2-Speciale","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25}},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"gpt-4-turbo-vision":{"id":"gpt-4-turbo-vision","name":"GPT-4 Turbo Vision","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30}},"cohere-embed-v-4-0":{"id":"cohere-embed-v-4-0","type":"embedding","name":"Embed v4","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":1536},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.12,"output":0}},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005}},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-07-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}}},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"model-router":{"id":"model-router","name":"Model Router","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"model-router","attachment":true,"reasoning":false,"tool_call":true,"release_date":"2025-05-19","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":16384},"provider":{"npm":"@ai-sdk/azure","shape":"completions"},"cost":{"input":0.14,"output":0}},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-chat-latest":{"id":"gpt-chat-latest","name":"GPT Chat Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-05-05","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"beta","cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5-instant"},"llama-4-maverick-17b-128e-instruct-fp8":{"id":"llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B 128E Instruct FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":16384},"provider":{"shape":"completions"},"cost":{"input":0.25,"output":1}},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"codestral-2501":{"id":"codestral-2501","name":"Codestral 25.01","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"provider":{"shape":"completions"},"cost":{"input":0.3,"output":0.9}},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.02,"output":0}},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-31","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"gpt-image-1":{"id":"gpt-image-1","type":"image","name":"GPT-Image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":40,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1"},"text-embedding-ada-002":{"id":"text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.1,"output":0}},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"phi-4-mini-reasoning":{"id":"phi-4-mini-reasoning","name":"Phi-4-mini-reasoning","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"phi-4-multimodal":{"id":"phi-4-multimodal","name":"Phi-4-multimodal","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"phi","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.08,"output":0.32,"input_audio":4}},"mistral-small-2503":{"id":"mistral-small-2503","name":"Mistral Small 3.1","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-03-01","last_updated":"2025-03-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.1,"output":0.3}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"gpt-3.5-turbo-0125":{"id":"gpt-3.5-turbo-0125","name":"GPT-3.5 Turbo 0125","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":0.5,"output":1.5}},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"cohere-command-a":{"id":"cohere-command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"provider":{"shape":"completions"},"cost":{"input":2.5,"output":10}},"gpt-image-2.5-sunburst":{"id":"gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"phi-4-reasoning":{"id":"phi-4-reasoning","name":"Phi-4-reasoning","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-08-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"cost":{"input":0.13,"output":0}},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek-V4-Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.19,"output":0.51},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-01-14","last_updated":"2026-01-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"grok-4-20-reasoning":{"id":"grok-4-20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"gpt-3.5-turbo-instruct":{"id":"gpt-3.5-turbo-instruct","name":"GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-09-21","last_updated":"2023-09-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":4096},"status":"deprecated","cost":{"input":1.5,"output":2}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"grok-4-20-non-reasoning":{"id":"grok-4-20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"phi-4-mini":{"id":"phi-4-mini","name":"Phi-4-mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"phi-4":{"id":"phi-4","name":"Phi-4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"codex-mini":{"id":"codex-mini","name":"Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-05-16","last_updated":"2025-05-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.5,"output":6,"cache_read":0.375}},"gpt-4-turbo":{"id":"gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"gpt-3.5-turbo-1106":{"id":"gpt-3.5-turbo-1106","name":"GPT-3.5 Turbo 1106","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":1,"output":2}},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"cohere-embed-v3-english":{"id":"cohere-embed-v3-english","type":"embedding","name":"Embed v3 English","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"claude-mythos-5":{"id":"claude-mythos-5","name":"Claude Mythos 5","description":"Restricted Claude model for advanced cybersecurity and biology research workflows","family":"claude-mythos","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_COGNITIVE_SERVICES_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-mythos-5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"mistral-medium-2505":{"id":"mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.4,"output":2}},"ministral-3b":{"id":"ministral-3b","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.04,"output":0.04}}}},"sambanova":{"id":"sambanova","env":["SAMBANOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.sambanova.ai/v1","name":"SambaNova","doc":"https://docs.sambanova.ai/cloud/docs/get-started/supported-models","models":{"DeepSeek-V3.1":{"id":"DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":7168},"cost":{"input":3,"output":4.5},"canonical_model_id":"deepseek/deepseek-v3.1"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.22,"output":0.59},"canonical_model_id":"openai/gpt-oss-120b"},"gemma-4-31B-it":{"id":"gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.38,"output":1.15},"canonical_model_id":"google/gemma-4-31b-it"},"Meta-Llama-3.3-70B-Instruct":{"id":"Meta-Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":3072},"cost":{"input":0.6,"output":1.2},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.6,"output":2.4,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"DeepSeek-V3.2":{"id":"DeepSeek-V3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":7168},"cost":{"input":3,"output":4.5},"canonical_model_id":"deepseek/deepseek-v3.2"}}},"infomaniak":{"id":"infomaniak","env":["INFOMANIAK_API_KEY","INFOMANIAK_PRODUCT_ID"],"npm":"@ai-sdk/openai-compatible","api":"https://api.infomaniak.com/2/ai/${INFOMANIAK_PRODUCT_ID}/openai/v1","name":"Infomaniak","doc":"https://www.infomaniak.com/en/hosting/ai-services/open-source-models","models":{"bge_multilingual_gemma2":{"id":"bge_multilingual_gemma2","type":"embedding","name":"BGE Multilingual Gemma2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-07-25","last_updated":"2026-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"input":8000,"output":3584},"cost":{"input":0.08,"output":0}},"mini_lm_l12_v2":{"id":"mini_lm_l12_v2","type":"embedding","name":"All-MiniLM-L12-v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2021-08-30","last_updated":"2021-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128,"input":128,"output":384},"cost":{"input":0,"output":0},"canonical_model_id":"sentence-transformers/all-minilm-l12-v2"},"mistralai/Mistral-Small-4-119B-2603":{"id":"mistralai/Mistral-Small-4-119B-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.25,"output":0.93},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/Ministral-3-14B-Instruct-2512":{"id":"mistralai/Ministral-3-14B-Instruct-2512","name":"Ministral 3 14B Instruct","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"input":100000,"output":25600},"status":"beta","cost":{"input":0.37,"output":0.5},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"input":100000,"output":32768},"cost":{"input":0.25,"output":0.5},"canonical_model_id":"google/gemma-4-31b-it"},"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8":{"id":"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8","name":"Nemotron 3 Nano 30B A3B FP8","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":262144},"status":"deprecated","cost":{"input":0.06,"output":0.25},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"status":"beta","cost":{"input":0.74,"output":3.72},"canonical_model_id":"moonshotai/kimi-k2.6"},"Qwen/Qwen3.5-397B-A17B-FP8":{"id":"Qwen/Qwen3.5-397B-A17B-FP8","name":"Qwen3.5 397B-A17B FP8","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65536},"status":"beta","cost":{"input":0.99,"output":4.46},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3.5-122B-A10B-FP8":{"id":"Qwen/Qwen3.5-122B-A10B-FP8","name":"Qwen3.5 122B-A10B FP8","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65536},"status":"deprecated","cost":{"input":0.5,"output":3.97},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"swiss-ai/Apertus-v1.5-70B":{"id":"swiss-ai/Apertus-v1.5-70B","name":"Apertus v1.5 70B","description":"Open, ethically-sourced Swiss AI model for multilingual, multimodal chat and instruction following","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-07-24","last_updated":"2026-08-01","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"input":100000,"output":8192},"status":"beta","cost":{"input":0.87,"output":3.1}}}},"databricks":{"id":"databricks","env":["DATABRICKS_HOST","DATABRICKS_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://${DATABRICKS_HOST}/ai-gateway/mlflow/v1","name":"Databricks","doc":"https://docs.databricks.com/aws/en/machine-learning/foundation-models/","models":{"databricks-gpt-5-nano":{"id":"databricks-gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"databricks-gpt-5-6-luna":{"id":"databricks-gpt-5-6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"databricks-gpt-5":{"id":"databricks-gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"databricks-gpt-5-4":{"id":"databricks-gpt-5-4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"databricks-gemini-3-flash":{"id":"databricks-gemini-3-flash","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":0.9,"output":5.4,"cache_read":0.09,"input_audio":1.8},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"databricks-kimi-k2-7-code":{"id":"databricks-kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"databricks-gpt-5-6-sol":{"id":"databricks-gpt-5-6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"databricks-claude-opus-4-6":{"id":"databricks-claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"databricks-gpt-5-6-terra":{"id":"databricks-gpt-5-6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"databricks-claude-opus-4-5":{"id":"databricks-claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"databricks-gpt-5-2":{"id":"databricks-gpt-5-2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"databricks-gemini-2-5-flash":{"id":"databricks-gemini-2-5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"databricks-gemini-3-1-flash-lite":{"id":"databricks-gemini-3-1-flash-lite","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":2.7,"cache_read":0.045,"input_audio":0.9},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"databricks-gpt-5-mini":{"id":"databricks-gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"databricks-glm-5-2":{"id":"databricks-glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"experimental":{"modes":{"fast":{"cost":{"input":2.45,"output":7.7,"cache_read":0.455},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"databricks-claude-sonnet-4-5":{"id":"databricks-claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"databricks-gpt-5-4-nano":{"id":"databricks-gpt-5-4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"databricks-gpt-oss-20b":{"id":"databricks-gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.07,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"databricks-gpt-5-1":{"id":"databricks-gpt-5-1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"databricks-gemini-3-pro":{"id":"databricks-gemini-3-pro","name":"Gemini 3 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3-pro-preview"},"databricks-claude-haiku-4-5":{"id":"databricks-claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"databricks-claude-sonnet-4":{"id":"databricks-claude-sonnet-4","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"databricks-gemini-2-5-pro":{"id":"databricks-gemini-2-5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"databricks-gpt-5-4-mini":{"id":"databricks-gpt-5-4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"databricks-claude-opus-4-1":{"id":"databricks-claude-opus-4-1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"databricks-gemini-3-5-flash":{"id":"databricks-gemini-3-5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":2.7,"output":16.2,"cache_read":0.27,"input_audio":2.7},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"databricks-gpt-oss-120b":{"id":"databricks-gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":25000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"databricks-claude-sonnet-4-6":{"id":"databricks-claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"databricks-gpt-5-5":{"id":"databricks-gpt-5-5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"databricks-claude-opus-4-7":{"id":"databricks-claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"databricks-gemini-3-1-pro":{"id":"databricks-gemini-3-1-pro","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"experimental":{"modes":{"fast":{"cost":{"input":3.6,"output":21.6,"cache_read":0.36,"input_audio":3.6},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"input_audio":2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"}}},"wandb":{"id":"wandb","env":["WANDB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inference.wandb.ai/v1","name":"CoreWeave","doc":"https://docs.wandb.ai/inference","models":{"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B","description":"Gemma 4 31B Dense is designed for advanced reasoning, agentic workflows, and longer context and is natively trained on 140+ languages.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.34,"cache_read":0.1},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B","description":"Gemma 4 26B A4B is a multimodal MoE model with LoRA support and function calling for agentic workflows.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B","name":"Nemotron 3.5 Super VL","description":"Nemotron 3.5 Super VL is a multimodal MoE model built for visual understanding, reasoning, and agentic workflows.","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B","name":"Nemotron 3.5 Lightning","description":"Nemotron 3.5 Lightning is an MoE model built for fast, reliable agentic tasks across use cases such as financial services, cybersecurity, telecom, and retail.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.07,"output":0.2,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B":{"id":"nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B","name":"Nemotron 3 Ultra","description":"Nemotron 3 Ultra is a powerful MoE model designed for long-running agents across coding, deep research, and enterprise automation.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.15,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax M3","description":"MiniMax M3 is a multimodal MoE model with 23B active parameters optimized for coding and agentic workflows.","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.23,"output":0.96,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"reflection/Beam-1.0-501B-A23B":{"id":"reflection/Beam-1.0-501B-A23B","name":"Beam 1.0 501B A23B","description":"Beam is a 501B-parameter MoE model with 23B active parameters built for coding, reasoning, and agentic workflows.","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1,"cache_read":0.04},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B","description":"Multilingual model excelling in conversational tasks, detailed instruction-following, and coding.","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.71,"output":0.71,"cache_read":0.71},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/Llama-3.1-8B-Instruct":{"id":"meta-llama/Llama-3.1-8B-Instruct","name":"Llama 3.1 8B","description":"Efficient conversational model optimized for responsive multilingual chatbot interactions.","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.22,"output":0.22,"cache_read":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Kimi K2.6 is a multimodal Mixture-of-Experts language model featuring 32 billion activated parameters and a total of 1 trillion parameters.","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.65,"output":3.41,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Kimi K2.7 Code is a 1T-parameter MoE model with 32B active parameters purpose-built for long-horizon agentic coding and software engineering.","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.71,"output":3.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Qwen3.8-27B is a dense multimodal model suited for coding, research, vision, and long-running agent tasks.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":3,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B A3B","description":"Qwen3.6-35B-A3B is an MoE multimodal model with 262K context optimized for agentic coding workflows.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":1.25,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B","description":"Granite 4.2 8B is an instruct model capable of enhanced tool calling, instruction following, and chat capabilities.","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.15,"cache_read":0.05},"canonical_model_id":"ibm/granite-4.2-8b"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"A large hybrid model that supports both thinking and non-thinking modes via prompt templates.","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":161000,"output":161000},"cost":{"input":0.55,"output":1.65,"cache_read":0.55},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash is a multimodal MoE model for coding, reasoning, and agentic workloads with long contexts.","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.65,"cache_read":0.002},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"DeepSeek V4-Flash-0731 is an MoE model great for coding, reasoning, and agentic workloads.","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.13,"output":0.28,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4-Pro-0813 is a 1.6T-parameter MoE model excelling at advanced reasoning, coding, and complex agentic workloads.","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.31,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"gpt-oss-120b","description":"Efficient Mixture-of-Experts model designed for high-reasoning, agentic and general-purpose use cases.","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.03,"output":0.17,"cache_read":0.03},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"gpt-oss-20b","description":"Lower latency Mixture-of-Experts model trained on OpenAI's Harmony response format with reasoning capabilities.","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.03,"output":0.13,"cache_read":0.03},"canonical_model_id":"openai/gpt-oss-20b"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM 5.3","description":"GLM-5.3 is a 744B-parameter MoE model built for advanced reasoning, software engineering, and long-horizon agentic tasks.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.2,"output":4,"cache_read":0.19},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"GLM-5.3-Flash is a natively multimodal model with 320B total parameters and 18B active parameters.","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM 5.2","description":"GLM-5.2 is a Mixture-of-Experts language model featuring 40 billion activated parameters and a total of 744 billion parameters.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.62,"output":2.42,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"}}},"merge-gateway":{"id":"merge-gateway","env":["MERGE_GATEWAY_API_KEY"],"npm":"merge-gateway-ai-sdk-provider","api":"https://api-gateway.merge.dev/v1/ai-sdk","name":"Merge Gateway","doc":"https://docs.merge.dev/merge-gateway","models":{"google/gemini-embedding-001":{"id":"google/gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":4096},"cost":{"input":0.15,"output":0},"canonical_model_id":"google/gemini-embedding-001"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.08,"output":0.45,"cache_read":0.04},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Gemini 3.1 Flash Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B It","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-2.5-computer-use-preview-10-2025":{"id":"google/gemini-2.5-computer-use-preview-10-2025","name":"Gemini 2.5 Computer Use Preview (10-2025)","description":"Specialized Gemini 2.5 model for browser-control agents that automate UI tasks","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-10-07","last_updated":"2025-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":1.25,"output":10},"canonical_model_id":"google/gemini-2.5-computer-use-preview-10-2025"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-flash-lite-latest"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.09,"output":0.29},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Gemini 2.5 Flash Image","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Gemini 3 Pro Image","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":2,"output":12},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.04,"output":0.08},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-flash-latest"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.13,"output":0.4},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-embedding-2":{"id":"google/gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.2,"output":0},"canonical_model_id":"google/gemini-embedding-2"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash-Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"nvidia/nemotron-3.5-lightning-30b-a3b":{"id":"nvidia/nemotron-3.5-lightning-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-nano-9b-v2":{"id":"nvidia/nemotron-nano-9b-v2","name":"Nemotron Nano 9B","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-09","release_date":"2025-08-18","last_updated":"2025-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.06,"output":0.23},"canonical_model_id":"nvidia/nemotron-nano-9b-v2"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"output":32000},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":2,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.56,"output":1.68,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.003625},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.035,"output":0.07,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-pro-0423":{"id":"deepseek/deepseek-v4-pro-0423","name":"DeepSeek V4 Pro 0423","description":"DeepSeek V4 Pro initial snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.65,"output":3.3},"canonical_model_id":"deepseek/deepseek-v4-pro-0423"},"deepseek/deepseek-v4-flash-0423":{"id":"deepseek/deepseek-v4-flash-0423","name":"DeepSeek V4 Flash 0423","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.139,"output":0.278},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 (0528)","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":32768},"cost":{"input":1.35,"output":5.4,"cache_read":0.945},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek/deepseek-v3":{"id":"deepseek/deepseek-v3","name":"DeepSeek V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":81920},"cost":{"input":0.27,"output":1.1},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek/deepseek-v4-flash-0731-fast":{"id":"deepseek/deepseek-v4-flash-0731-fast","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.28,"output":0.4,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006,"cache_write":0.1125},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-r1":{"id":"deepseek/deepseek-r1","name":"DeepSeek R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":40960},"cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"meta/llama-3.3-70b-instruct":{"id":"meta/llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.22,"output":0.5,"cache_read":0.11},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta/llama-3.1-8b-instruct":{"id":"meta/llama-3.1-8b-instruct","name":"Llama 3.1 8B","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta/llama-3.1-70b-instruct":{"id":"meta/llama-3.1-70b-instruct","name":"Llama 3.1 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":2048},"cost":{"input":0.99,"output":0.99},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"openai/whisper-1":{"id":"openai/whisper-1","type":"transcription","name":"Whisper","description":"General-purpose speech recognition model for transcription and translation","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2023-03-01","last_updated":"2023-03-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":448000,"output":4096},"provider":{"npm":"@ai-sdk/openai","api":"https://api-gateway.merge.dev/v1"},"canonical_model_id":"openai/whisper-1"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT-OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.05,"output":0.25},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-oss-safeguard-120b":{"id":"openai/gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT-OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.04,"output":0.15,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-audio-mini":{"id":"openai/gpt-audio-mini","type":"chat","name":"GPT Audio Mini","description":"Cost-efficient audio model for speech input and output with the Chat Completions API","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.6,"output":2.4},"canonical_model_id":"openai/gpt-audio-mini"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-audio":{"id":"openai/gpt-audio","type":"chat","name":"GPT Audio","description":"Audio model for speech input and output with the Chat Completions API","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-audio"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.2-chat-latest":{"id":"openai/gpt-5.2-chat-latest","name":"GPT-5.2 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3 Mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":5,"output":15},"canonical_model_id":"openai/gpt-4o-2024-05-13"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5-search-api":{"id":"openai/gpt-5-search-api","type":"chat","name":"GPT-5 Search API","description":"GPT-5 model trained for web search in the Chat Completions API","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5-search-api"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4 Mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.07,"output":0.2,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-5.1-chat-latest":{"id":"openai/gpt-5.1-chat-latest","name":"GPT-5.1 Chat Latest","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":30,"output":60},"canonical_model_id":"openai/gpt-4"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o Mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.3-chat-latest":{"id":"openai/gpt-5.3-chat-latest","name":"GPT-5.3 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5-chat-latest":{"id":"openai/gpt-5-chat-latest","name":"GPT-5 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-chat-latest"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"Glm 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.7-flash":{"id":"zai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM-4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":50000}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.05,"output":3.3,"cache_read":0.195},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM-4.7 FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.7,"output":2.2,"cache_read":0.13},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM-5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM-5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5 (20251101)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4-1-20250805":{"id":"anthropic/claude-opus-4-1-20250805","name":"Claude Opus 4.1 (20250805)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-4-5-20250929":{"id":"anthropic/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5 (20250929)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":128000}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-haiku-4-5-20251001":{"id":"anthropic/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5 (20251001)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/sakana-namazu":{"id":"sakana/sakana-namazu","name":"Sakana Namazu","description":"Japanese-specialized reasoning model based on Kimi K2.6 and tuned for Japanese language, culture, and business workflows","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 Highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":128000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax M2.5 Highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"mistral/devstral-medium-latest":{"id":"mistral/devstral-medium-latest","name":"Devstral 2 (latest)","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/devstral-medium-latest"},"mistral/devstral-2512":{"id":"mistral/devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/devstral-2512"},"mistral/magistral-medium-latest":{"id":"mistral/magistral-medium-latest","name":"Magistral Medium (latest)","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral-medium","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-03-17","last_updated":"2025-03-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2,"output":5},"canonical_model_id":"mistral/magistral-medium-latest"},"mistral/pixtral-large-latest":{"id":"mistral/pixtral-large-latest","name":"Pixtral Large (latest)","description":"Mistral's larger vision model for document-heavy image understanding and chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2024-11-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/pixtral-large-latest"},"mistral/mistral-small-latest":{"id":"mistral/mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-latest"},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"mistral/mistral-large-latest"},"mistral/mistral-medium-latest":{"id":"mistral/mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-latest"},"mistral/codestral-latest":{"id":"mistral/codestral-latest","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.3,"output":0.9},"canonical_model_id":"mistral/codestral-latest"},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/mistral-medium-2505":{"id":"mistral/mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":0.4,"output":2,"cache_read":0.04},"canonical_model_id":"mistral/mistral-medium-2505"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":0.13},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3-VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.21,"output":1.9,"cache_read":0.105},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3-vl-30b-a3b-instruct":{"id":"qwen/qwen3-vl-30b-a3b-instruct","name":"Qwen3-VL 30B-A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6,"cache_read":0.0216},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.115,"output":0.287,"cache_read":0.023,"cache_write":0.14375},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.4126,"output":2.4754},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3-vl-plus":{"id":"qwen/qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.143,"output":1.434,"cache_read":0.0286,"cache_write":0.17875},"canonical_model_id":"alibaba/qwen3-vl-plus"},"qwen/qwen3.8-2.4t-a95b":{"id":"qwen/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":1010000},"cost":{"input":2.5,"output":6.25,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"qwen/qwen3-max-preview":{"id":"qwen/qwen3-max-preview","name":"Qwen3 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-05","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1.08,"output":4.8},"canonical_model_id":"alibaba/qwen3-max-preview"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.15,"output":0.8,"cache_read":0.075},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":977000,"output":128000},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3-vl-flash":{"id":"qwen/qwen3-vl-flash","name":"Qwen3-VL Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.022,"output":0.215,"cache_read":0.0044,"cache_write":0.0275},"canonical_model_id":"alibaba/qwen3-vl-flash"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.276,"output":1.651,"cache_read":0.0276,"cache_write":0.345},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":1.1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3-235b-a22b-instruct-2507":{"id":"qwen/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.0574},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.029,"output":0.287,"cache_read":0.0029,"cache_write":0.03625},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.115,"output":0.688,"cache_read":0.0115,"cache_write":0.14375},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.172,"output":1.032,"cache_read":0.0344},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.115,"output":0.917,"cache_read":0.023},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.5-4b":{"id":"qwen/qwen3.5-4b","name":"Qwen3.5 4B","description":"Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.04,"output":0.07},"canonical_model_id":"alibaba/qwen3.5-4b"},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.165,"output":0.99,"cache_read":0.0165,"cache_write":0.20625},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3-coder-480b-a35b-instruct":{"id":"qwen/qwen3-coder-480b-a35b-instruct","name":"Qwen3 Coder 480B A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.22,"output":1.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.359,"output":1.434,"cache_read":0.0718,"cache_write":0.44875},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.574,"output":2.294,"cache_read":0.1148,"cache_write":0.7175},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.825,"output":2.4755,"cache_read":0.165,"cache_write":1.03125},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen-flash":{"id":"qwen/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.022,"output":0.216,"cache_read":0.0044,"cache_write":0.0275},"canonical_model_id":"alibaba/qwen-flash"},"qwen/qwen3-235b-a22b":{"id":"qwen/qwen3-235b-a22b","name":"Qwen3 235B A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.287,"output":1.147,"cache_read":0.0574},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.248,"output":1.485,"cache_read":0.0496},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.108,"output":1.076,"cache_read":0.0216},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen3.6-max-preview":{"id":"qwen/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":1.31,"output":7.88},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.057,"output":0.459,"cache_read":0.020357},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-vl-235b-a22b-thinking":{"id":"qwen/qwen3-vl-235b-a22b-thinking","name":"Qwen3-VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.287,"output":2.867,"cache_read":0.0574},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.086,"output":0.688,"cache_read":0.0172},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.144,"output":0.574,"cache_read":0.0288,"cache_write":0.18},"canonical_model_id":"alibaba/qwen3-coder-flash"},"moonshot/kimi-k2.7-code":{"id":"moonshot/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot/kimi-k3.1-pro":{"id":"moonshot/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshot/kimi-k2.5":{"id":"moonshot/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshot/kimi-k3":{"id":"moonshot/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshot/kimi-k2.6":{"id":"moonshot/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshot/kimi-k2.7-code-highspeed":{"id":"moonshot/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"bytedance/dola-seed-2.0-code-preview":{"id":"bytedance/dola-seed-2.0-code-preview","name":"Dola Seed 2.0 Code (preview)","description":"Preview coding model for repository understanding, refactors, and engineering tasks","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-03-28","last_updated":"2026-03-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3}},"bytedance/dola-seed-2.0-lite":{"id":"bytedance/dola-seed-2.0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.25,"output":2},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance/dola-seed-2.0-mini":{"id":"bytedance/dola-seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance/dola-seed-2.0-code":{"id":"bytedance/dola-seed-2.0-code","name":"Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4,"output":2.4},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"bytedance/dola-seed-2.0-pro":{"id":"bytedance/dola-seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"writer/palmyra-x4":{"id":"writer/palmyra-x4","name":"Palmyra X4","description":"Enterprise language model for workflow automation, coding, data analysis, and tool use","family":"palmyra","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2024-10-09","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32000},"cost":{"input":2.5,"output":10},"canonical_model_id":"writer/palmyra-x4"},"writer/palmyra-x5":{"id":"writer/palmyra-x5","name":"Palmyra X5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"palmyra","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":250000},"cost":{"input":0.6,"output":6},"canonical_model_id":"writer/palmyra-x5"},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B 12-2024","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+ 08-2024","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R 08-2024","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"cohere/command-a-03-2025":{"id":"cohere/command-a-03-2025","name":"Command A 03-2025","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-03-2025"}}},"nan":{"id":"nan","env":["NAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.nan.builders/v1","name":"NaN","doc":"https://nan.builders/docs/models","models":{"glm5.3-flash":{"id":"glm5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"qwen3.6":{"id":"qwen3.6","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"glm5.3":{"id":"glm5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"},"qwen-image-2.1":{"id":"qwen-image-2.1","type":"image","name":"Qwen-Image-2.1","description":"Unified text-to-image generation and image editing model with native RGBA transparency and reference-image editing","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-14","last_updated":"2026-09-21","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":8192,"output":0},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen-image-2.1"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gemma4":{"id":"gemma4","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-26b-a4b-it"}}},"cline-pass":{"id":"cline-pass","env":["CLINE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.cline.bot/api/v1","name":"ClinePass","doc":"https://docs.cline.bot/getting-started/clinepass","models":{"cline-pass/muse-spark-1.3-contributor":{"id":"cline-pass/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.002},"canonical_model_id":"meta/muse-spark-1.3"},"cline-pass/mimo-v2.6-pro":{"id":"cline-pass/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"cline-pass/kimi-k2.7-code":{"id":"cline-pass/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"cline-pass/glm-5.2":{"id":"cline-pass/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"cline-pass/qwen3.7-plus":{"id":"cline-pass/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.12,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"},"cline-pass/deepseek-v4.1-flash":{"id":"cline-pass/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"cline-pass/deepseek-v4-pro":{"id":"cline-pass/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"cline-pass/minimax-m3":{"id":"cline-pass/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"cline-pass/mimo-v2.5":{"id":"cline-pass/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"cline-pass/kimi-k3":{"id":"cline-pass/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"cline-pass/kimi-k2.6":{"id":"cline-pass/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"cline-pass/mimo-v2.6-flash":{"id":"cline-pass/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"cline-pass/glm-5.3":{"id":"cline-pass/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"cline-pass/qwen3.7-max":{"id":"cline-pass/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"cline-pass/mimo-v2.5-pro":{"id":"cline-pass/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.74,"output":3.48,"cache_read":0.0145},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"cline-pass/qwen3.8-max":{"id":"cline-pass/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"cline-pass/deepseek-v4-flash":{"id":"cline-pass/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"cline-pass/glm-5.3-flash":{"id":"cline-pass/glm-5.3-flash","name":"cline-pass/glm-5.3-flash","description":"Latest natively multimodal model in the GLM-5 series","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"openai":{"id":"openai","env":["OPENAI_API_KEY"],"npm":"@ai-sdk/openai","name":"OpenAI","doc":"https://platform.openai.com/docs/models","models":{"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"gpt-image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0},"canonical_model_id":"openai/gpt-image-1.5"},"o1":{"id":"o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5}},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":0.7,"output":2.8,"cache_read":0.175},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.4,"output":1.6,"cache_read":0.1}},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":0.2,"output":0.8,"cache_read":0.05},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.1,"output":0.4,"cache_read":0.025}},"gpt-image-1-mini":{"id":"gpt-image-1-mini","type":"image","name":"gpt-image-1-mini","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0}},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.625,"output":5,"cache_read":0.0625},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"o1-pro":{"id":"o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":150,"output":600}},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}}},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":20,"output":100,"cache_read":2,"cache_write":25},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}},"ultrafast":{"cost":{"input":60,"output":300,"cache_read":6,"cache_write":75},"provider":{"body":{"service_tier":"ultrafast"}}}}},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-5.3-codex-spark":{"id":"gpt-5.3-codex-spark","name":"GPT-5.3 Codex Spark","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-spark","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":100000,"output":32000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5}},"chatgpt-image-latest":{"id":"chatgpt-image-latest","type":"image","name":"chatgpt-image-latest","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0}},"gpt-4o-2024-08-06":{"id":"gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25}},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.125,"output":1,"cache_read":0.0125},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"gpt-daybreak-blue-latest":{"id":"gpt-daybreak-blue-latest","name":"Daybreak Blue","description":"Alias for flagship general-purpose models with safeguards for defensive cybersecurity work","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-5.2-pro":{"id":"gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168}},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}}},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.2,"output":1,"cache_read":0.02,"cache_write":0.25},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"gpt-realtime-2.1":{"id":"gpt-realtime-2.1","type":"realtime","name":"GPT-Realtime-2.1","description":"Realtime speech-to-speech model with configurable reasoning, tool use, and robust voice-agent behavior","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","audio","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"input":96000,"output":32000},"cost":{"input":4,"output":24,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-2.1"},"gpt-4o-2024-11-20":{"id":"gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25}},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"flex":{"cost":{"input":0.025,"output":0.2,"cache_read":0.0025},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.05,"output":0.4,"cache_read":0.005}},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.375,"output":2.25,"cache_read":0.0375},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075}},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}}},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"gpt-4o-2024-05-13":{"id":"gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":8.75,"output":26.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":15}},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.875,"output":7,"cache_read":0.0875},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-01","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":1536},"cost":{"input":0.02,"output":0}},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}},"ultrafast":{"cost":{"input":12,"output":60,"cache_read":0.6,"cache_write":15},"provider":{"body":{"service_tier":"ultrafast"}}}}},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}}},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120}},"gpt-image-1":{"id":"gpt-image-1","type":"image","name":"gpt-image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0},"status":"deprecated","canonical_model_id":"openai/gpt-image-1"},"text-embedding-ada-002":{"id":"text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2022-12","release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.1,"output":0}},"gpt-5.6":{"id":"gpt-5.6","name":"GPT-5.6","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":2,"output":8,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.55,"output":2.2,"cache_read":0.1375},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.1,"output":4.4,"cache_read":0.275}},"gpt-3.5-turbo":{"id":"gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"status":"deprecated","cost":{"input":0.5,"output":1.5,"cache_read":0}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"flex":{"cost":{"input":0.1,"output":0.625,"cache_read":0.01},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.2,"output":1.25,"cache_read":0.02}},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":4.25,"output":17,"cache_read":2.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":10,"cache_read":1.25}},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":1,"output":4,"cache_read":0.25},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5}},"gpt-4":{"id":"gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"status":"deprecated","cost":{"input":30,"output":60}},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-01","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":3072},"cost":{"input":0.13,"output":0}},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":0.25,"output":1,"cache_read":0.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.6,"cache_read":0.075}},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"gpt-image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"input":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.625,"output":5,"cache_read":0.0625},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"o3-pro":{"id":"o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80}},"gpt-4-turbo":{"id":"gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":10,"output":30}},"gpt-daybreak-red-latest":{"id":"gpt-daybreak-red-latest","name":"Daybreak Red","description":"Alias for advanced cybersecurity models for authorized vulnerability research and security testing","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":12.5,"output":75,"cache_read":1.25,"cache_write":15.625},"canonical_model_id":"openai/gpt-5.6-cyber"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}},"pro":{"provider":{"body":{"reasoning":{"mode":"pro"}}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"}}},"stepfun-ai":{"id":"stepfun-ai","env":["STEPFUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stepfun.ai/v1","name":"StepFun (Global)","doc":"https://platform.stepfun.ai/docs/en/overview/concept","models":{"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"input":1024000,"output":65536},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"step-3.5-flash-2603":{"id":"step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"stepaudio-2.5-asr":{"id":"stepaudio-2.5-asr","type":"transcription","name":"StepAudio 2.5 ASR","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"step","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-24","last_updated":"2026-07-02","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"stepaudio-2.5-tts":{"id":"stepaudio-2.5-tts","type":"speech","name":"StepAudio 2.5 TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-16","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}},"step-3.5-flash":{"id":"step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"step-tts-2":{"id":"step-tts-2","type":"speech","name":"Step TTS 2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"step","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-01","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/openai"}}}},"abliteration-ai":{"id":"abliteration-ai","env":["ABLIT_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.abliteration.ai/v1","name":"abliteration.ai","doc":"https://docs.abliteration.ai/models","models":{"abliterated-model-large":{"id":"abliterated-model-large","name":"Abliterated Model Large","description":"GLM-5.2 model abliterated and finetuned for cyber, ML red teaming, and agent testing","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-25","last_updated":"2026-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":5,"output":5,"cache_read":0.5}},"abliterated-model-large-v2":{"id":"abliterated-model-large-v2","name":"Abliterated Model Large V2","description":"GLM-5.3 model abliterated and finetuned for cyber, ML red teaming, and agent testing","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-29","last_updated":"2026-08-31","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":5,"output":5,"cache_read":0.5}},"abliterated-model":{"id":"abliterated-model","name":"Abliterated Model","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-01-06","last_updated":"2026-07-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":150000,"input":150000,"output":8192},"cost":{"input":3,"output":3,"cache_read":0.3}}}},"lynkr":{"id":"lynkr","env":["LYNKR_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"http://127.0.0.1:8081/v1","name":"Lynkr","doc":"https://github.com/Fast-Editor/Lynkr","models":{"lynkr-auto":{"id":"lynkr-auto","name":"Lynkr Auto (complexity routing)","description":"Virtual model: Lynkr scores each request on complexity and routes it to the tier model the user configured (local Ollama/llama.cpp for simple requests, configured cloud providers for complex ones).","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-03","last_updated":"2026-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0}}}},"anyapi":{"id":"anyapi","env":["ANYAPI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.anyapi.ai/v1","name":"AnyAPI","doc":"https://docs.anyapi.ai","models":{"mistralai/devstral-2512":{"id":"mistralai/devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","canonical_model_id":"mistral/devstral-2512"},"mistralai/mistral-large-2512":{"id":"mistralai/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"canonical_model_id":"mistral/mistral-large-2512"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3-pro-preview":{"id":"google/gemini-3-pro-preview","name":"Gemini 3 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3-pro-preview"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-2.5-flash-lite"},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"canonical_model_id":"xai/grok-4.3"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-chat"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-r1":{"id":"deepseek/deepseek-r1","name":"DeepSeek Reasoner","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"canonical_model_id":"deepseek/deepseek-reasoner"},"perplexity/sonar-reasoning-pro":{"id":"perplexity/sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"canonical_model_id":"perplexity/sonar-pro"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5-mini"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5.2"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o4-mini"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"canonical_model_id":"openai/o3"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"canonical_model_id":"openai/gpt-5"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":127999}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"canonical_model_id":"anthropic/claude-opus-4-7"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"canonical_model_id":"cohere/command-r-plus-08-2024"}}},"zai":{"id":"zai","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.z.ai/api/paas/v4","name":"Z.AI","doc":"https://docs.z.ai/guides/overview/pricing","models":{"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5v-turbo"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5v"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.2,"output":1.1,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-air"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-4.5-flash":{"id":"glm-4.5-flash","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"glm-5.3-flashx":{"id":"glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-4.6v-flash":{"id":"glm-4.6v-flash","name":"GLM-4.6V-Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"glm-4.7-flashx":{"id":"glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.07,"output":0.4,"cache_read":0.01,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.05,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v"},"glm-5-turbo":{"id":"glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5-turbo"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"poolside":{"id":"poolside","env":["POOLSIDE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.poolside.ai/v1","name":"Poolside","doc":"https://platform.poolside.ai","models":{"poolside/laguna-m.1":{"id":"poolside/laguna-m.1","name":"Laguna M.1","description":"Poolside's open-weight model for agentic coding and long-horizon work","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"poolside/laguna-m.1"},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"poolside/laguna-xs-2.1"},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"poolside/laguna-s-2.1"}}},"vancine":{"id":"vancine","env":["VANCINE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://vancine.com/v1","name":"Vancine","doc":"https://vancine.com/docs","models":{"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.24,"output":0.96,"cache_read":0.0048},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"status":"deprecated","cost":{"input":0.24,"output":0.96,"cache_read":0.048},"canonical_model_id":"minimax/MiniMax-M3"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.67,"output":2,"cache_read":0.034},"canonical_model_id":"tencent/hy4-preview"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.12,"output":0.38,"cache_read":0.013},"canonical_model_id":"alibaba/qwen3.8-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.4,"output":12,"cache_read":0.24},"canonical_model_id":"moonshotai/kimi-k3"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.12,"output":3.52,"cache_read":0.208},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.6,"output":4.8,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-max"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.12,"output":0.4,"cache_read":0.024},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"submodel":{"id":"submodel","env":["SUBMODEL_INSTAGEN_ACCESS_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://llm.submodel.ai/v1","name":"submodel","doc":"https://submodel.gitbook.io","models":{"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.2,"output":0.3},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen3 235B A22B Thinking 2507","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.2,"output":0.6},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8","name":"Qwen3 Coder 480B A35B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":75000,"output":163840},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":75000,"output":163840},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":75000,"output":163840},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"deepseek/deepseek-r1-0528"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"openai/gpt-oss-120b"},"zai-org/GLM-4.5-FP8":{"id":"zai-org/GLM-4.5-FP8","name":"GLM 4.5 FP8","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.2,"output":0.8},"canonical_model_id":"zhipuai/glm-4.5"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"GLM 4.5 Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"zhipuai/glm-4.5-air"}}},"azure":{"id":"azure","env":["AZURE_RESOURCE_NAME","AZURE_API_KEY"],"npm":"@ai-sdk/azure","name":"Azure","doc":"https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/models","models":{"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"GPT-Image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":32,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1.5"},"o1":{"id":"o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5}},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.71,"output":0.71}},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}},"gpt-image-2.5-flare":{"id":"gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"phi-4-reasoning-plus":{"id":"phi-4-reasoning-plus","name":"Phi-4-reasoning-plus","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek-V4.1-Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"llama-4-scout-17b-16e-instruct":{"id":"llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B 16E Instruct","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.78}},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek-V4-Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":1.74,"output":3.48},"canonical_model_id":"deepseek/deepseek-v4-pro"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"status":"beta","provider":{"shape":"completions"},"canonical_model_id":"xai/grok-4.7"},"cohere-embed-v3-multilingual":{"id":"cohere-embed-v3-multilingual","type":"embedding","name":"Embed v3 Multilingual","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"deepseek-v3.2-speciale":{"id":"deepseek-v3.2-speciale","name":"DeepSeek-V3.2-Speciale","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25}},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"gpt-4-turbo-vision":{"id":"gpt-4-turbo-vision","name":"GPT-4 Turbo Vision","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30}},"cohere-embed-v-4-0":{"id":"cohere-embed-v-4-0","type":"embedding","name":"Embed v4","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":1536},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.12,"output":0}},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005}},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-07-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}}},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"model-router":{"id":"model-router","name":"Model Router","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"model-router","attachment":true,"reasoning":false,"tool_call":true,"release_date":"2025-05-19","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":16384},"provider":{"npm":"@ai-sdk/azure","shape":"completions"},"cost":{"input":0.14,"output":0}},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-chat-latest":{"id":"gpt-chat-latest","name":"GPT Chat Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-05-05","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"beta","cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5-instant"},"llama-4-maverick-17b-128e-instruct-fp8":{"id":"llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B 128E Instruct FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":16384},"provider":{"shape":"completions"},"cost":{"input":0.25,"output":1}},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":28,"cache_read":0.35},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"codestral-2501":{"id":"codestral-2501","name":"Codestral 25.01","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"provider":{"shape":"completions"},"cost":{"input":0.3,"output":0.9}},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.02,"output":0}},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-31","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5-pro":{"id":"gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"gpt-image-1":{"id":"gpt-image-1","type":"image","name":"GPT-Image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":40,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1"},"text-embedding-ada-002":{"id":"text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.1,"output":0}},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"phi-4-mini-reasoning":{"id":"phi-4-mini-reasoning","name":"Phi-4-mini-reasoning","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"phi-4-multimodal":{"id":"phi-4-multimodal","name":"Phi-4-multimodal","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"phi","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.08,"output":0.32,"input_audio":4}},"mistral-small-2503":{"id":"mistral-small-2503","name":"Mistral Small 3.1","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-03-01","last_updated":"2025-03-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"provider":{"shape":"completions"},"cost":{"input":0.1,"output":0.3}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-14","last_updated":"2025-11-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025}},"gpt-3.5-turbo-0125":{"id":"gpt-3.5-turbo-0125","name":"GPT-3.5 Turbo 0125","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":0.5,"output":1.5}},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-27","last_updated":"2025-06-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.2,"output":0.5,"cache_read":0.05}},"cohere-command-a":{"id":"cohere-command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"provider":{"shape":"completions"},"cost":{"input":2.5,"output":10}},"gpt-image-2.5-sunburst":{"id":"gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"phi-4-reasoning":{"id":"phi-4-reasoning","name":"Phi-4-reasoning","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"phi","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-08-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"cost":{"input":0.13,"output":0}},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.58,"output":1.68}},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek-V4-Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.19,"output":0.51},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-01-14","last_updated":"2026-01-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175}},"grok-4-20-reasoning":{"id":"grok-4-20-reasoning","name":"Grok 4.20 (Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"gpt-3.5-turbo-instruct":{"id":"gpt-3.5-turbo-instruct","name":"GPT-3.5 Turbo Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-09-21","last_updated":"2023-09-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":4096},"status":"deprecated","cost":{"input":1.5,"output":2}},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"grok-4-20-non-reasoning":{"id":"grok-4-20-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-04-08","last_updated":"2026-04-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262000,"output":8192},"status":"beta","provider":{"shape":"completions"},"cost":{"input":2,"output":6}},"phi-4-mini":{"id":"phi-4-mini","name":"Phi-4-mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"phi","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.075,"output":0.3}},"phi-4":{"id":"phi-4","name":"Phi-4","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"provider":{"shape":"completions"},"cost":{"input":0.125,"output":0.5}},"codex-mini":{"id":"codex-mini","name":"Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-05-16","last_updated":"2025-05-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.5,"output":6,"cache_read":0.375}},"gpt-4-turbo":{"id":"gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"gpt-3.5-turbo-1106":{"id":"gpt-3.5-turbo-1106","name":"GPT-3.5 Turbo 1106","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-08","release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":4096},"status":"deprecated","provider":{"shape":"completions"},"cost":{"input":1,"output":2}},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"cohere-embed-v3-english":{"id":"cohere-embed-v3-english","type":"embedding","name":"Embed v3 English","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-07","last_updated":"2023-11-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/models"},"cost":{"input":0.1,"output":0}},"claude-mythos-5":{"id":"claude-mythos-5","name":"Claude Mythos 5","description":"Restricted Claude model for advanced cybersecurity and biology research workflows","family":"claude-mythos","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"beta","provider":{"npm":"@ai-sdk/anthropic","api":"https://${AZURE_RESOURCE_NAME}.services.ai.azure.com/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-mythos-5"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"mistral-medium-2505":{"id":"mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"provider":{"shape":"completions"},"cost":{"input":0.4,"output":2}},"ministral-3b":{"id":"ministral-3b","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"provider":{"shape":"completions"},"cost":{"input":0.04,"output":0.04}}}},"watsonx":{"id":"watsonx","env":["WATSONX_AI_APIKEY","WATSONX_AI_PROJECT_ID"],"npm":"watsonx-ai-provider","name":"watsonx.ai","doc":"https://www.ibm.com/docs/en/watsonx/saas?topic=solutions-supported-foundation-models","models":{"mistralai/mistral-small-3-1-24b-instruct-2503":{"id":"mistralai/mistral-small-3-1-24b-instruct-2503","name":"Mistral Small 3.1 24B","description":"Efficient multimodal model for instruction following, coding, reasoning, and function calling","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.106,"output":0.318},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"meta-llama/llama-3-3-70b-instruct":{"id":"meta-llama/llama-3-3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.7526,"output":0.7526},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-4-maverick-17b-128e-instruct-fp8":{"id":"meta-llama/llama-4-maverick-17b-128e-instruct-fp8","name":"Llama 4 Maverick 17B 128E Instruct FP8","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.371,"output":1.484},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.159,"output":0.636},"canonical_model_id":"openai/gpt-oss-120b"},"ibm/granite-4-h-small":{"id":"ibm/granite-4-h-small","name":"Granite-4.0-H-Small","description":"Open-weight hybrid model for enterprise chat, coding, retrieval-augmented generation, and tool-calling workloads","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0636,"output":0.265},"canonical_model_id":"ibm/granite-4-h-small"}}},"minimax-cn-coding-plan":{"id":"minimax-cn-coding-plan","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.cn/anthropic/v1","name":"MiniMax Token Plan (minimax.cn)","doc":"https://platform.minimaxi.com/docs/token-plan/intro","models":{"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M3.1-Flash-Preview":{"id":"MiniMax-M3.1-Flash-Preview","name":"MiniMax-M3.1-Flash-Preview","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-27","last_updated":"2026-09-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M3.1-Flash-Preview"},"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}}}},"friendli":{"id":"friendli","env":["FRIENDLI_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://api.friendli.ai/serverless/v1","name":"Friendli","doc":"https://friendli.ai/docs/guides/serverless_endpoints/introduction","models":{"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.26,"output":3.96,"cache_read":0.234},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"evroc":{"id":"evroc","env":["EVROC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://models.think.evroc.com/v1","name":"evroc","doc":"https://docs.evroc.com/products/think/overview.html","models":{"mistralai/Mistral-Medium-3.5-128B":{"id":"mistralai/Mistral-Medium-3.5-128B","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.725,"output":6.9},"canonical_model_id":"mistral/mistral-medium-2604"},"mistralai/Voxtral-Small-24B-2507":{"id":"mistralai/Voxtral-Small-24B-2507","type":"transcription","name":"Voxtral Small 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"release_date":"2025-03-01","last_updated":"2025-03-01","modalities":{"input":["audio","text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"provider":{"npm":"@ai-sdk/openai"}},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.144,"output":0.575},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"nvidia/Llama-3.3-70B-Instruct-FP8":{"id":"nvidia/Llama-3.3-70B-Instruct-FP8","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":1.15,"output":1.15},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.4375,"output":5.75},"canonical_model_id":"moonshotai/kimi-k2.6"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8-27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.87,"output":3.5},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.345,"output":1.38},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-Reranker-4B":{"id":"Qwen/Qwen3-Reranker-4B","type":"reranking","name":"Qwen3 Reranker 4B","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.0575,"output":0}},"Qwen/Qwen3-Embedding-8B":{"id":"Qwen/Qwen3-Embedding-8B","type":"embedding","name":"Qwen3 Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":4096},"cost":{"input":0.115,"output":0.115},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.23,"output":0.92},"canonical_model_id":"openai/gpt-oss-120b"},"openai/whisper-large-v3-turbo":{"id":"openai/whisper-large-v3-turbo","type":"transcription","name":"Whisper Large v3 Turbo","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.0023,"output":0.0023,"output_audio":2.3},"canonical_model_id":"openai/whisper-large-v3-turbo"},"openai/whisper-large-v3":{"id":"openai/whisper-large-v3","type":"transcription","name":"Whisper 3 Large","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.0023,"output":0.0023,"output_audio":2.3},"canonical_model_id":"openai/whisper-large-v3"},"intfloat/multilingual-e5-large-instruct":{"id":"intfloat/multilingual-e5-large-instruct","type":"embedding","name":"E5 Multi-Lingual Large Embeddings 0.6B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"cost":{"input":0.114,"output":0.114}},"evroc/roc":{"id":"evroc/roc","name":"roc","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-06-06","last_updated":"2026-06-06","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":2.875,"output":11.516}},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":131072},"cost":{"input":1.4375,"output":5.75},"canonical_model_id":"zhipuai/glm-5.2"},"KBLab/kb-whisper-large":{"id":"KBLab/kb-whisper-large","type":"transcription","name":"KB Whisper","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"provider":{"npm":"@ai-sdk/openai"}}}},"synthetic":{"id":"synthetic","env":["SYNTHETIC_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.synthetic.new/openai/v1","name":"Synthetic","doc":"https://synthetic.new/pricing","models":{"hf:deepseek-ai/DeepSeek-V4.1-Flash":{"id":"hf:deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.6,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"hf:deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"hf:deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.6,"output":1.2,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"hf:openai/gpt-oss-120b":{"id":"hf:openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.1,"output":0.1,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-120b"},"hf:MiniMaxAI/MiniMax-M3":{"id":"hf:MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"status":"deprecated","cost":{"input":0.6,"output":1.2,"cache_read":0.6},"canonical_model_id":"minimax/MiniMax-M3"},"hf:nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4":{"id":"hf:nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.3,"output":1,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"hf:zai-org/GLM-5.3":{"id":"hf:zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"hf:zai-org/GLM-4.7-Flash":{"id":"hf:zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":65536},"cost":{"input":0.1,"output":0.5,"cache_read":0.02},"canonical_model_id":"zhipuai/glm-4.7-flash"},"hf:zai-org/GLM-5.3-Flash":{"id":"hf:zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":0.15,"output":0.5,"cache_read":0.04},"canonical_model_id":"zhipuai/glm-5.3-flash"},"hf:zai-org/GLM-5.2":{"id":"hf:zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.2"},"hf:moonshotai/Kimi-K2.7-Code":{"id":"hf:moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.95},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"hf:moonshotai/Kimi-K3":{"id":"hf:moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":65536},"cost":{"input":3,"output":15,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"hf:Qwen/Qwen3.6-27B":{"id":"hf:Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.45,"output":3.6,"cache_read":0.45},"canonical_model_id":"alibaba/qwen3.6-27b"}}},"freemodel":{"id":"freemodel","env":["FREEMODEL_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://cc.freemodel.dev/v1","name":"FreeModel","doc":"https://freemodel.dev","models":{"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5},"canonical_model_id":"openai/gpt-5.5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075,"cache_write":0.75},"canonical_model_id":"openai/gpt-5.4-mini"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.freemodel.dev/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175,"cache_write":1.75},"canonical_model_id":"openai/gpt-5.3-codex"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"}}},"alibaba-token-plan-cn":{"id":"alibaba-token-plan-cn","env":["ALIBABA_TOKEN_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan.cn-beijing.maas.aliyuncs.com/compatible-mode/v1","name":"Alibaba Token Plan (China)","doc":"https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview","models":{"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"happyhorse-1.1-r2v":{"id":"happyhorse-1.1-r2v","type":"video","name":"HappyHorse 1.1 Reference-to-Video","description":"Video model for reference-guided video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.cn-beijing.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"qwen3.8-max-preview":{"id":"qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"qwen-image-2.0-pro":{"id":"qwen-image-2.0-pro","type":"image","name":"Qwen Image 2.0 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen-image-2.0":{"id":"qwen-image-2.0","type":"image","name":"Qwen Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"happyhorse-1.1-i2v":{"id":"happyhorse-1.1-i2v","type":"video","name":"HappyHorse 1.1 Image-to-Video","description":"Video model for image-to-video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.cn-beijing.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-plus"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"wan2.7-image-pro":{"id":"wan2.7-image-pro","type":"image","name":"Wan2.7 Image Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-flash"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"happyhorse-1.1-t2v":{"id":"happyhorse-1.1-t2v","type":"video","name":"HappyHorse 1.1 Text-to-Video","description":"Video model for prompt-driven text-to-video generation","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.cn-beijing.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-max"},"wan2.7-image":{"id":"wan2.7-image","type":"image","name":"Wan2.7 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.cn-beijing.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-03","last_updated":"2025-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0,"output":0}},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"input":196601,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"vercel":{"id":"vercel","env":["AI_GATEWAY_API_KEY"],"npm":"@ai-sdk/gateway","name":"Vercel AI Gateway","doc":"https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway","models":{"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember-1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"google/gemini-embedding-001":{"id":"google/gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"google/gemini-embedding-001"},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3.1-flash-image"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Gemini multimodal model for text, image, audio, video, and document tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/text-embedding-005":{"id":"google/text-embedding-005","type":"embedding","name":"Text Embedding 005","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-01","last_updated":"2024-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/veo-3.1-fast-generate-001":{"id":"google/veo-3.1-fast-generate-001","type":"video","name":"Veo 3.1 Fast Generate","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-06-08","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.5-transcribe-live":{"id":"google/gemini-3.5-transcribe-live","type":"transcription","name":"Gemini 3.5 Transcribe Live","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"google/gemini-3.5-transcribe-live"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/veo-3.1-generate-001":{"id":"google/veo-3.1-generate-001","type":"video","name":"Veo 3.1","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-06-08","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite)","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana (Gemini 2.5 Flash Image)","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash-image"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-3.8-flash-lite-tts":{"id":"google/gemini-3.8-flash-lite-tts","type":"speech","name":"Gemini 3.8 Flash-Lite TTS","description":"Fast, cost-efficient text-to-speech model for high-volume narration and voice agent pipelines","family":"gemini-flash-lite","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.5,"output":6,"output_audio":6},"canonical_model_id":"google/gemini-3.8-flash-lite-tts"},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.8-live-extended-thinking":{"id":"google/gemini-3.8-live-extended-thinking","type":"realtime","name":"Gemini 3.8 Live Extended Thinking","description":"Live API voice model with configurable background reasoning for multi-step real-time agents","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-09-15","last_updated":"2026-09-15","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.75,"output":4.5,"input_audio":3,"output_audio":12},"canonical_model_id":"google/gemini-3.8-live-extended-thinking"},"google/gemini-3.8-live":{"id":"google/gemini-3.8-live","type":"realtime","name":"Gemini 3.8 Live","description":"Low-latency Live API model for real-time voice agents with interleaved reasoning and async tools","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-09-15","last_updated":"2026-09-15","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.75,"output":4.5,"input_audio":3,"output_audio":12},"canonical_model_id":"google/gemini-3.8-live"},"google/text-multilingual-embedding-002":{"id":"google/text-multilingual-embedding-002","type":"embedding","name":"Text Multilingual Embedding 002","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-01","last_updated":"2024-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-omni-flash-preview":{"id":"google/gemini-omni-flash-preview","type":"video","name":"Gemini Omni Flash Preview","description":"Omni-modal model for text, vision, audio, and multimodal agent tasks","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":57920},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-omni-flash-preview"},"google/veo-3.0-generate-001":{"id":"google/veo-3.0-generate-001","type":"video","name":"Veo 3.0","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-20","last_updated":"2026-06-08","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-3.5-transcribe":{"id":"google/gemini-3.5-transcribe","type":"transcription","name":"Gemini 3.5 Transcribe","description":"Speech-to-text model with language detection, speaker diarization, word timestamps, and custom vocabulary","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":98304,"output":32768},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":2,"output":12,"input_audio":2},"canonical_model_id":"google/gemini-3.5-transcribe"},"google/gemini-3-flash":{"id":"google/gemini-3-flash","name":"Gemini 3 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.5,"output":3,"cache_read":0.05}},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-embedding-2":{"id":"google/gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"google/gemini-embedding-2"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/veo-3.1-lite-generate-001":{"id":"google/veo-3.1-lite-generate-001","type":"video","name":"Veo 3.1 Lite Generate","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"google/gemini-3.8-flash-tts":{"id":"google/gemini-3.8-flash-tts","type":"speech","name":"Gemini 3.8 Flash TTS","description":"Flagship text-to-speech model for expressive, studio-grade voices and long-form multi-speaker audio","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.5,"output":9,"output_audio":9},"canonical_model_id":"google/gemini-3.8-flash-tts"},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Gemini Nano Banana 2.1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-nano-banana-2.1"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Gemini 3.1 Flash Image Preview (Nano Banana 2)","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"google/veo-3.0-fast-generate-001":{"id":"google/veo-3.0-fast-generate-001","type":"video","name":"Veo 3.0 Fast Generate","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-07-31","last_updated":"2026-06-08","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/nemotron-nano-12b-v2-vl":{"id":"nvidia/nemotron-nano-12b-v2-vl","name":"Nvidia Nemotron Nano 12B V2 VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.2,"output":0.6},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"nvidia/nemotron-3.5-lightning-free":{"id":"nvidia/nemotron-3.5-lightning-free","name":"Nemotron 3.5 Lightning Free","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"NVIDIA Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.15,"output":0.65},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65000},"cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-nano-9b-v2":{"id":"nvidia/nemotron-nano-9b-v2","name":"Nvidia Nemotron Nano 9B V2","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-08-18","last_updated":"2025-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.06,"output":0.23},"canonical_model_id":"nvidia/nemotron-nano-9b-v2"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nemotron 3.5 Lightning 30B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"NVIDIA Nemotron 3.5 Super VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.2,"output":0.5,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"inception/mercury-2":{"id":"inception/mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536},"cost":{"input":0.04,"output":0.15,"cache_read":0.004},"canonical_model_id":"inception/mercury-2.5"},"inception/mercury-coder-small":{"id":"inception/mercury-coder-small","name":"Mercury Coder Small Beta","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"mercury","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":16384},"cost":{"input":0.25,"output":1}},"inception/mercury-3":{"id":"inception/mercury-3","name":"Mercury 3","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01}},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo M2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131100},"cost":{"input":0.119,"output":0.238,"cache_read":0.0025},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1050000,"output":131000},"cost":{"input":0.3045,"output":0.609,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"recraft/recraft-v4.1-pro":{"id":"recraft/recraft-v4.1-pro","type":"image","name":"Recraft V4.1 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v3":{"id":"recraft/recraft-v3","type":"image","name":"Recraft V3","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-30","last_updated":"2024-10","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1-utility":{"id":"recraft/recraft-v4.1-utility","type":"image","name":"Recraft V4.1 Utility","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4":{"id":"recraft/recraft-v4","type":"image","name":"Recraft V4","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4-pro":{"id":"recraft/recraft-v4-pro","type":"image","name":"Recraft V4 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v2":{"id":"recraft/recraft-v2","type":"image","name":"Recraft V2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-13","last_updated":"2024-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1-utility-pro":{"id":"recraft/recraft-v4.1-utility-pro","type":"image","name":"Recraft V4.1 Utility Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1":{"id":"recraft/recraft-v4.1","type":"image","name":"Recraft V4.1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"recraft/recraft-v4.1-flash":{"id":"recraft/recraft-v4.1-flash","type":"image","name":"Recraft V4.1 Flash","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"recraft","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"elorian/glimpse":{"id":"elorian/glimpse","name":"Glimpse","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":220000,"output":4096},"cost":{"input":0.6,"output":3,"cache_read":0.06}},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"meituan/longcat-2.5-preview":{"id":"meituan/longcat-2.5-preview","name":"LongCat 2.5 Preview","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.5-preview"},"topaz/proteus":{"id":"topaz/proteus","type":"video","name":"Proteus","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"topaz/starlight-precise-2.6":{"id":"topaz/starlight-precise-2.6","type":"video","name":"Starlight Precise 2.6","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"topaz/wonder-3.5":{"id":"topaz/wonder-3.5","type":"image","name":"Wonder 3.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"reflection/beam":{"id":"reflection/beam","name":"Beam 1.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"reflection/beam-free":{"id":"reflection/beam-free","name":"Beam 1.0 (Free)","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"fish-audio/s2.1-pro":{"id":"fish-audio/s2.1-pro","type":"speech","name":"S2.1 Pro","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-28","last_updated":"2026-07-28","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/s1":{"id":"fish-audio/s1","type":"speech","name":"S1","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-20","last_updated":"2025-10-20","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/s2-pro":{"id":"fish-audio/s2-pro","type":"speech","name":"S2 Pro","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/drama-3":{"id":"fish-audio/drama-3","type":"speech","name":"Drama 3","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"fish-audio/transcribe-1":{"id":"fish-audio/transcribe-1","type":"transcription","name":"Transcribe-1","description":"Speech transcription model for accurate audio-to-text and captioning workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-01","last_updated":"2026-03-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"tencent/hy-mt2-plus":{"id":"tencent/hy-mt2-plus","name":"Tencent Hy-MT2-Plus","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":4000},"cost":{"input":0.074,"output":0.295}},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Tencent Hy4 Preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy-mt2-pro":{"id":"tencent/hy-mt2-pro","name":"Tencent Hy-MT2-Pro","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":4000},"cost":{"input":0.074,"output":0.295}},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"tencent/hy-mt2-lite":{"id":"tencent/hy-mt2-lite","name":"Tencent Hy-MT2-Lite","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":4000},"cost":{"input":0.044,"output":0.177}},"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262100,"output":80000},"cost":{"input":0.25,"output":0.8,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"deepseek/deepseek-v4.1-flash-1017-fast":{"id":"deepseek/deepseek-v4.1-flash-1017-fast","name":"DeepSeek V4.1 Flash 1017 Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.075,"output":0.3,"cache_read":0.0015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.075,"output":0.3,"cache_read":0.0015},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek-V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":128000},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.066},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.076,"output":0.153,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v3.2-thinking":{"id":"deepseek/deepseek-v3.2-thinking","name":"DeepSeek V3.2 Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":0.62,"output":1.85},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4.1-flash-fast":{"id":"deepseek/deepseek-v4.1-flash-fast","name":"DeepSeek V4.1 Flash Fast","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.6,"output":2.4,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":0.62,"output":1.85},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.0068},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek/deepseek-r1":{"id":"deepseek/deepseek-r1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":1.35,"output":5.4},"canonical_model_id":"deepseek/deepseek-r1"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k3-fast":{"id":"moonshotai/kimi-k3-fast","name":"Kimi K3 Fast","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k3.1-pro-highspeed":{"id":"moonshotai/kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4}},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code High Speed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"meta/llama-4-maverick":{"id":"meta/llama-4-maverick","name":"Llama-4-Maverick-17B-128E-Instruct-FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.24,"output":0.97},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/llama-3.3-70b":{"id":"meta/llama-3.3-70b","name":"Llama-3.3-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/llama-3.1-8b":{"id":"meta/llama-3.1-8b","name":"Llama 3.1 8B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.22,"output":0.22},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/llama-4-scout":{"id":"meta/llama-4-scout","name":"Llama-4-Scout-17B-16E-Instruct-FP8","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.17,"output":0.66},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-image-1.0":{"id":"meta/muse-image-1.0","type":"image","name":"Muse Image 1.0","description":"Agentic image generation and editing model that plans, searches and refines before rendering","family":"muse","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-07","last_updated":"2026-07-07","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"meta/muse-image-1.0"},"meta/llama-3.1-70b":{"id":"meta/llama-3.1-70b","name":"Llama 3.1 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.72,"output":0.72},"canonical_model_id":"meta/llama-3.1-70b-instruct"},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Open Llama multimodal model for image understanding and text reasoning","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"voyage/voyage-3.5-lite":{"id":"voyage/voyage-3.5-lite","type":"embedding","name":"voyage-3.5-lite","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-4-large":{"id":"voyage/voyage-4-large","type":"embedding","name":"voyage-4-large","description":"General-purpose and multilingual embedding model optimized for retrieval quality","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/voyage-4-large"},"voyage/voyage-code-3":{"id":"voyage/voyage-code-3","type":"embedding","name":"voyage-code-3","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-12-04","last_updated":"2024-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-3-large":{"id":"voyage/voyage-3-large","type":"embedding","name":"voyage-3-large","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-01-07","last_updated":"2024-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-law-2":{"id":"voyage/voyage-law-2","type":"embedding","name":"voyage-law-2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-04-15","last_updated":"2024-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/rerank-2.5-lite":{"id":"voyage/rerank-2.5-lite","type":"reranking","name":"Voyage Rerank 2.5 Lite","description":"Instruction-following reranker optimized for latency and quality","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-2.5-lite"},"voyage/rerank-3":{"id":"voyage/rerank-3","type":"reranking","name":"Voyage Rerank 3","description":"Reranker optimized for retrieval quality, successor to rerank-2.5","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-3"},"voyage/voyage-4-lite":{"id":"voyage/voyage-4-lite","type":"embedding","name":"voyage-4-lite","description":"Lightweight general-purpose embedding model optimized for latency and cost","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/voyage-4-lite"},"voyage/voyage-finance-2":{"id":"voyage/voyage-finance-2","type":"embedding","name":"voyage-finance-2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-06-03","last_updated":"2024-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/rerank-2.5":{"id":"voyage/rerank-2.5","type":"reranking","name":"Voyage Rerank 2.5","description":"Instruction-following reranker optimized for retrieval quality","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-2.5"},"voyage/voyage-code-2":{"id":"voyage/voyage-code-2","type":"embedding","name":"voyage-code-2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-01","last_updated":"2024-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-3.5":{"id":"voyage/voyage-3.5","type":"embedding","name":"voyage-3.5","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"voyage/voyage-4":{"id":"voyage/voyage-4","type":"embedding","name":"voyage-4","description":"General-purpose and multilingual embedding model balancing retrieval quality and cost","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/voyage-4"},"voyage/rerank-3-lite":{"id":"voyage/rerank-3-lite","type":"reranking","name":"Voyage Rerank 3 Lite","description":"Reranker optimized for latency and cost, successor to rerank-2.5-lite","family":"voyage","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"voyageai/rerank-3-lite"},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Sonar search model for current answers, retrieval, and citation-backed chat","family":"sonar","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-02","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":8000},"cost":{"input":0.25,"output":2.5,"cache_write":0.0625}},"perplexity/pplx-embed-v1-0.6b":{"id":"perplexity/pplx-embed-v1-0.6b","type":"embedding","name":"Embed v1 0.6b","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"v0","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"perplexity/pplx-embed-v1-4b":{"id":"perplexity/pplx-embed-v1-4b","type":"embedding","name":"Embed v1 4b","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"spacexai/grok-4.20-non-reasoning-beta":{"id":"spacexai/grok-4.20-non-reasoning-beta","name":"Grok 4.20 Beta Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.4,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-4.20-reasoning-beta":{"id":"spacexai/grok-4.20-reasoning-beta","name":"Grok 4.20 Beta Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-stt":{"id":"spacexai/grok-stt","type":"transcription","name":"Grok STT","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"spacexai/grok-voice-think-fast-2.0":{"id":"spacexai/grok-voice-think-fast-2.0","type":"realtime","name":"Grok Voice Think Fast 2.0","description":"Realtime voice model for low-latency spoken conversations","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"spacexai/grok-4.20-non-reasoning":{"id":"spacexai/grok-4.20-non-reasoning","name":"Grok 4.20 Non-Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"spacexai/grok-4.7":{"id":"spacexai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"spacexai/grok-imagine-image":{"id":"spacexai/grok-imagine-image","type":"image","name":"Grok Imagine Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-image"},"spacexai/grok-4.6":{"id":"spacexai/grok-4.6","name":"Grok 4.6","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"spacexai/grok-4.20-multi-agent":{"id":"spacexai/grok-4.20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"release_date":"2026-03-10","last_updated":"2026-03-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-4.20-multi-agent-beta":{"id":"spacexai/grok-4.20-multi-agent-beta","name":"Grok 4.20 Multi Agent Beta","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}}},"spacexai/grok-imagine-video":{"id":"spacexai/grok-imagine-video","type":"video","name":"Grok Imagine","description":"Video model for text-to-video, image-to-video, video editing and extension with native synchronized audio","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-video"},"spacexai/grok-4.3":{"id":"spacexai/grok-4.3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"spacexai/grok-4.20-reasoning":{"id":"spacexai/grok-4.20-reasoning","name":"Grok 4.20 Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"spacexai/grok-build-0.1":{"id":"spacexai/grok-build-0.1","name":"Grok Build 0.1","description":"Grok coding model for agentic engineering, edits, and codebase workflows","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"spacexai/grok-imagine-image-2.0":{"id":"spacexai/grok-imagine-image-2.0","type":"image","name":"Grok Imagine Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":64000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-image-2.0"},"spacexai/grok-tts":{"id":"spacexai/grok-tts","type":"speech","name":"Grok TTS","description":"Convert text into spoken audio with a single API call. The API supports a rich set of expressive voices, inline speech tags for fine-grained delivery control, and output formats from high-fidelity MP3 to telephony-optimized μ-law.","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":15000,"output":15000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-tts"},"spacexai/grok-imagine-video-1.5-lite":{"id":"spacexai/grok-imagine-video-1.5-lite","type":"video","name":"Grok Imagine Video 1.5 Lite","description":"Cost-efficient video model for text-to-video and image-to-video generation","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-video-1.5-lite"},"spacexai/grok-imagine-video-1.5":{"id":"spacexai/grok-imagine-video-1.5","type":"video","name":"Grok Imagine Video 1.5","description":"Video model for text-to-video, image-to-video, and reference-to-video generation with optional voice references","family":"grok","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-30","last_updated":"2026-05-30","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"xai/grok-imagine-video-1.5"},"spacexai/grok-4.5":{"id":"spacexai/grok-4.5","name":"Grok 4.5","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"alibaba/wan-v2.7-r2v":{"id":"alibaba/wan-v2.7-r2v","type":"video","name":"Wan v2.7 Reference-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.7-plus":{"id":"alibaba/qwen3.7-plus","name":"Qwen 3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"},"alibaba/wan-v3.0-video-prime":{"id":"alibaba/wan-v3.0-video-prime","type":"video","name":"Wan v3.0 Video Prime","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3-coder":{"id":"alibaba/qwen3-coder","name":"Qwen3 Coder 480B A35B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.22,"output":1.8,"cache_read":0.022,"tiers":[{"input":2.7,"output":13.5,"cache_read":0.54,"tier":{"type":"context","size":32001}},{"input":4.5,"output":22.5,"cache_read":0.9,"tier":{"type":"context","size":128001}}]},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"alibaba/wan-v2.5-t2v-preview":{"id":"alibaba/wan-v2.5-t2v-preview","type":"video","name":"Wan v2.5 Text-to-Video Preview","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-24","last_updated":"2025-09-24","modalities":{"input":["text","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.8-max-prime":{"id":"alibaba/qwen3.8-max-prime","name":"Qwen 3.8 Max Prime","description":"High-throughput edition of Qwen3.8 Max for coding, professional work, multimodal understanding, and long-running agent workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5,"cache_write":5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"alibaba/qwen-3-14b":{"id":"alibaba/qwen-3-14b","name":"Qwen3-14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.24}},"alibaba/qwen-3-235b":{"id":"alibaba/qwen-3-235b","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.22,"output":0.88},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"alibaba/qwen3.6-27b":{"id":"alibaba/qwen3.6-27b","name":"Qwen 3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.6-27b"},"alibaba/qwen3.8-2.4t-a95b":{"id":"alibaba/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"alibaba/qwen-3-30b":{"id":"alibaba/qwen-3-30b","name":"Qwen3-30B-A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"alibaba/qwen3-coder-next":{"id":"alibaba/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.5,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"alibaba/qwen3.8-flash":{"id":"alibaba/qwen3.8-flash","name":"Qwen 3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":128000},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"alibaba/qwen3.8-omni-flash":{"id":"alibaba/qwen3.8-omni-flash","name":"Qwen 3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"alibaba/qwen3.6-plus":{"id":"alibaba/qwen3.6-plus","name":"Qwen 3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":131072}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.2,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"alibaba/qwen3-next-80b-a3b-instruct":{"id":"alibaba/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262114,"output":262114},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"alibaba/qwen3-embedding-8b":{"id":"alibaba/qwen3-embedding-8b","type":"embedding","name":"Qwen3 Embedding 8B","description":"Multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"alibaba/qwen3-embedding-0.6b":{"id":"alibaba/qwen3-embedding-0.6b","type":"embedding","name":"Qwen3 Embedding 0.6B","description":"Compact multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"alibaba/qwen3-embedding-0.6b"},"alibaba/wan-v2.7-t2v":{"id":"alibaba/wan-v2.7-t2v","type":"video","name":"Wan v2.7 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.5-flash":{"id":"alibaba/qwen3.5-flash","name":"Qwen 3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"alibaba/qwen3.5-plus":{"id":"alibaba/qwen3.5-plus","name":"Qwen 3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":256001}}],"context_over_200k":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625}},"canonical_model_id":"alibaba/qwen3.5-plus"},"alibaba/wan-v2.6-i2v":{"id":"alibaba/wan-v2.6-i2v","type":"video","name":"Wan v2.6 Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/wan-v2.6-r2v":{"id":"alibaba/wan-v2.6-r2v","type":"video","name":"Wan v2.6 Reference-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.8-max-0902":{"id":"alibaba/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"alibaba/qwen-3-32b":{"id":"alibaba/qwen-3-32b","name":"Qwen 3.32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-32b"},"alibaba/qwen3-max-thinking":{"id":"alibaba/qwen3-max-thinking","name":"Qwen 3 Max Thinking","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-23","last_updated":"2025-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":1.2,"output":6,"cache_read":0.24,"tiers":[{"input":2.4,"output":12,"cache_read":0.48,"tier":{"type":"context","size":32001}},{"input":3,"output":15,"cache_read":0.6,"tier":{"type":"context","size":128001}}]}},"alibaba/wan-v2.6-t2v":{"id":"alibaba/wan-v2.6-t2v","type":"video","name":"Wan v2.6 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/wan-v2.6-r2v-flash":{"id":"alibaba/wan-v2.6-r2v-flash","type":"video","name":"Wan v2.6 Reference-to-Video Flash","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.7-max":{"id":"alibaba/qwen3.7-max","name":"Qwen 3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":64000},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"alibaba/qwen3-next-80b-a3b-thinking":{"id":"alibaba/qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B Thinking","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":1.2},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"alibaba/qwen3-vl-instruct":{"id":"alibaba/qwen3-vl-instruct","name":"Qwen3 VL Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.88,"cache_read":0.11}},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen 3.8 Max","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"alibaba/wan-v3.0-video":{"id":"alibaba/wan-v3.0-video","type":"video","name":"Wan v3.0 Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-23","last_updated":"2026-08-23","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3.8-27b":{"id":"alibaba/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.8-27b"},"alibaba/qwen3-embedding-4b":{"id":"alibaba/qwen3-embedding-4b","type":"embedding","name":"Qwen3 Embedding 4B","description":"Multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"alibaba/qwen3-embedding-4b"},"alibaba/qwen3.7-flash":{"id":"alibaba/qwen3.7-flash","name":"Qwen 3.7 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":64000},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"alibaba/wan-v2.6-i2v-flash":{"id":"alibaba/wan-v2.6-i2v-flash","type":"video","name":"Wan v2.6 Image-to-Video Flash","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"alibaba/qwen3-coder-30b-a3b":{"id":"alibaba/qwen3-coder-30b-a3b","name":"Qwen 3 Coder 30B A3B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"stealth/glyph-cluster":{"id":"stealth/glyph-cluster","name":"Glyph Cluster","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0,"output":0}},"liquid/lfm2.5-24b-a2b":{"id":"liquid/lfm2.5-24b-a2b","name":"LFM2.5-24B-A2B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8200},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"quiverai/arrow-2-telos":{"id":"quiverai/arrow-2-telos","name":"Arrow 2 Telos","description":"High-fidelity SVG generation model for complex vector work and long-context refinement","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"release_date":"2026-09-16","last_updated":"2026-09-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":6,"output":30,"cache_read":0.6,"cache_write":7.5},"canonical_model_id":"quiverai/arrow-2-telos"},"quiverai/arrow-2":{"id":"quiverai/arrow-2","name":"Arrow 2","description":"Fast SVG generation model for creation, vectorization, editing, and animation","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"release_date":"2026-09-16","last_updated":"2026-09-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5},"canonical_model_id":"quiverai/arrow-2"},"mixedbread/toast-1":{"id":"mixedbread/toast-1","name":"Toast 1","description":"Specialized search model for knowledge-intensive questions, multi-step retrieval, and evidence synthesis","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131000,"output":4000},"cost":{"input":0.3,"output":0.72,"cache_read":0.036},"canonical_model_id":"mixedbread/toast-1"},"klingai/kling-v2.6-t2v":{"id":"klingai/kling-v2.6-t2v","type":"video","name":"Kling v2.6 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.6-motion-control":{"id":"klingai/kling-v2.6-motion-control","type":"video","name":"Kling v2.6 Motion Control","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-21","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v3.0-i2v":{"id":"klingai/kling-v3.0-i2v","type":"video","name":"Kling v3.0 Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.5-turbo-i2v":{"id":"klingai/kling-v2.5-turbo-i2v","type":"video","name":"Kling v2.5 Turbo Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.5-turbo-t2v":{"id":"klingai/kling-v2.5-turbo-t2v","type":"video","name":"Kling v2.5 Turbo Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v2.6-i2v":{"id":"klingai/kling-v2.6-i2v","type":"video","name":"Kling v2.6 Image-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-12-21","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v3.0-motion-control":{"id":"klingai/kling-v3.0-motion-control","type":"video","name":"Kling v3.0 Motion Control","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-04","last_updated":"2026-03-04","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"klingai/kling-v3.0-t2v":{"id":"klingai/kling-v3.0-t2v","type":"video","name":"Kling v3.0 Text-to-Video","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ling","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"amazon/titan-embed-text-v2":{"id":"amazon/titan-embed-text-v2","type":"embedding","name":"Titan Text Embeddings V2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"titan-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-04-30","last_updated":"2024-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"amazon/nova-2-lite":{"id":"amazon/nova-2-lite","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.0825},"canonical_model_id":"amazon/nova-2-lite"},"amazon/nova-micro":{"id":"amazon/nova-micro","name":"Nova Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.035,"output":0.14,"cache_read":0.00875,"cache_write":0.035},"canonical_model_id":"amazon/nova-micro"},"amazon/nova-lite":{"id":"amazon/nova-lite","name":"Nova Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.06,"output":0.24,"cache_read":0.015,"cache_write":0.06},"canonical_model_id":"amazon/nova-lite"},"amazon/nova-pro":{"id":"amazon/nova-pro","name":"Nova Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.8,"output":3.2,"cache_read":0.2,"cache_write":0.8},"canonical_model_id":"amazon/nova-pro"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash-fin":{"id":"inclusionai/ling-3.0-flash-fin","name":"Ling 3.0 Flash Fin","description":"Finance-enhanced model for financial research, multi-step investment workflows, and long-horizon planning and execution","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-09-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-fin"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":0.021,"output":0.063,"cache_read":0.0042},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.06},"canonical_model_id":"inclusionai/ling-3.1-flash"},"inclusionai/ling-3.0-flash-sante":{"id":"inclusionai/ling-3.0-flash-sante","name":"Ling 3.0 Flash Sante","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash-sante"},"openai/gpt-image-1.5":{"id":"openai/gpt-image-1.5","type":"image","name":"GPT Image 1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":32,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1.5"},"openai/whisper-1":{"id":"openai/whisper-1","type":"transcription","name":"Whisper","description":"General-purpose speech recognition model for transcription and translation","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-03-01","last_updated":"2023-03-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/whisper-1"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"status":"deprecated","cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-realtime-1.5":{"id":"openai/gpt-realtime-1.5","type":"realtime","name":"GPT-Realtime-1.5","description":"Realtime speech-to-speech model for voice agents and customer support","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32000,"output":4096},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":4,"output":16,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-1.5"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":0.7,"output":2.8,"cache_read":0.175},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-live-1":{"id":"openai/gpt-live-1","type":"realtime","name":"GPT-Live 1","description":"Full-duplex voice model for natural conversations that delegates reasoning and tools to a backend agent","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/gpt-live-1"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":0.2,"output":0.8,"cache_read":0.06},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.1,"output":0.4,"cache_read":0.03},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-image-1-mini":{"id":"openai/gpt-image-1-mini","type":"image","name":"GPT Image 1 Mini","description":"Cost-efficient GPT Image 1 variant for image generation and editing","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":2,"output":8,"cache_read":0.2},"canonical_model_id":"openai/gpt-image-1-mini"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.1,"output":0.5,"cache_read":0.1},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-safeguard-120b":{"id":"openai/gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":112000,"output":16000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"openai/gpt-4.1-fast":{"id":"openai/gpt-4.1-fast","name":"GPT-4.1 (Fast)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"cost":{"input":3.5,"output":14,"cache_read":0.875},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-image-2.5-flare":{"id":"openai/gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT 5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT 5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.6-luna-fast":{"id":"openai/gpt-5.6-luna-fast","name":"GPT 5.6 Luna (Fast)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.4,"output":2.4,"cache_read":0.04,"cache_write":0.5,"tiers":[{"input":0.8,"output":3.6,"cache_read":0.08,"cache_write":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.8,"output":3.6,"cache_read":0.08,"cache_write":1}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1-Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":122880,"output":8192},"cost":{"input":0.03,"output":0.14},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.125,"output":1,"cache_read":0.0125},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.1-thinking":{"id":"openai/gpt-5.1-thinking","name":"GPT 5.1 Thinking","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-11-12","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125}},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT 5.2 ","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.1-thinking-fast":{"id":"openai/gpt-5.1-thinking-fast","name":"GPT 5.1 Thinking (Fast)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"release_date":"2025-11-12","last_updated":"2025-11-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2.5,"output":20,"cache_read":0.25}},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT 5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":872000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-realtime-mini":{"id":"openai/gpt-realtime-mini","type":"realtime","name":"GPT-Realtime mini","description":"Cost-efficient realtime speech-to-speech model for voice agents","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-10-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32000,"output":4096},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"input_audio":10,"output_audio":20},"canonical_model_id":"openai/gpt-realtime-mini"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-realtime-2.1":{"id":"openai/gpt-realtime-2.1","type":"realtime","name":"gpt-realtime-2.1","description":"Realtime speech-to-speech model with configurable reasoning, tool use, and robust voice-agent behavior","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"input":96000,"output":32000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":4,"output":24,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-2.1"},"openai/gpt-5-fast":{"id":"openai/gpt-5-fast","name":"GPT-5 (Fast)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":2.5,"output":20,"cache_read":0.25},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"flex":{"cost":{"input":0.025,"output":0.2,"cache_read":0.0025},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT 5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT 5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":872000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT 5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"status":"deprecated","cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.4-mini-fast":{"id":"openai/gpt-5.4-mini-fast","name":"GPT 5.4 Mini (Fast)","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.2-fast":{"id":"openai/gpt-5.2-fast","name":"GPT 5.2 (Fast)","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":3.5,"output":28,"cache_read":0.35},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.3-codex-fast":{"id":"openai/gpt-5.3-codex-fast","name":"GPT 5.3 Codex (Fast)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":3.5,"output":28,"cache_read":0.35},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-realtime-whisper":{"id":"openai/gpt-realtime-whisper","type":"transcription","name":"gpt-realtime-whisper","description":"Streaming speech-to-text model for low-latency transcript deltas from live audio","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/gpt-realtime-whisper"},"openai/gpt-4o-fast":{"id":"openai/gpt-4o-fast","name":"GPT-4o (Fast)","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"cost":{"input":4.25,"output":17,"cache_read":2.125},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-5.4-fast":{"id":"openai/gpt-5.4-fast","name":"GPT 5.4 (Fast)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT 5.6 Terra","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/text-embedding-3-small":{"id":"openai/text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Small, efficient OpenAI text embedding model","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/text-embedding-3-small"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-realtime-2":{"id":"openai/gpt-realtime-2","type":"realtime","name":"gpt-realtime-2","description":"Realtime speech-to-speech model with configurable reasoning and reliable tool use for voice agents","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"output":32000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":4,"output":24,"cache_read":0.4,"input_audio":32,"output_audio":64},"canonical_model_id":"openai/gpt-realtime-2"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT 5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":128000,"output":272000},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-4o-transcribe":{"id":"openai/gpt-4o-transcribe","type":"transcription","name":"GPT-4o Transcribe","description":"Speech-to-text model powered by GPT-4o for accurate transcription","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-06-01","release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":2.5,"output":10,"input_audio":2.5},"canonical_model_id":"openai/gpt-4o-transcribe"},"openai/text-embedding-ada-002":{"id":"openai/text-embedding-ada-002","type":"embedding","name":"text-embedding-ada-002","description":"Legacy OpenAI text embedding model","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2022-12-15","last_updated":"2022-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/text-embedding-ada-002"},"openai/o3-fast":{"id":"openai/o3-fast","name":"o3 (Fast)","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"cost":{"input":3.5,"output":14,"cache_read":0.875},"canonical_model_id":"openai/o3"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":2,"output":8,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.55,"output":2.2,"cache_read":0.1375},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"gpt-oss-safeguard-20b","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":112000,"output":16000},"cost":{"input":0.07,"output":0.2},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT 5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.03},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":4.25,"output":17,"cache_read":2.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/gpt-5.6-terra-fast":{"id":"openai/gpt-5.6-terra-fast","name":"GPT 5.6 Terra (Fast)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":24,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":36,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":36,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-image-2.5-sunburst":{"id":"openai/gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"openai/gpt-6-sol-fast":{"id":"openai/gpt-6-sol-fast","name":"GPT-6 Sol (Fast)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.5-fast":{"id":"openai/gpt-5.5-fast","name":"GPT 5.5 (Fast)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":872000,"output":128000},"cost":{"input":12.5,"output":75,"cache_read":1.25},"canonical_model_id":"openai/gpt-5.5"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"experimental":{"modes":{"fast":{"cost":{"input":3.5,"output":14,"cache_read":0.875},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":1,"output":4,"cache_read":0.25},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-4.1-mini-fast":{"id":"openai/gpt-4.1-mini-fast","name":"GPT-4.1 mini (Fast)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1014808,"output":32768},"cost":{"input":0.7,"output":2.8,"cache_read":0.175},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4o-mini-transcribe":{"id":"openai/gpt-4o-mini-transcribe","type":"transcription","name":"GPT-4o mini Transcribe","description":"Speech-to-text model powered by GPT-4o Mini for fast, accurate transcription","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2024-06-01","release_date":"2025-03-20","last_updated":"2025-03-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":1.25,"output":5,"input_audio":1.25},"canonical_model_id":"openai/gpt-4o-mini-transcribe"},"openai/text-embedding-3-large":{"id":"openai/text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Most capable OpenAI embedding model for English and non-English text","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/text-embedding-3-large"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT 5.6 Luna","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"experimental":{"modes":{"fast":{"cost":{"input":0.25,"output":1,"cache_read":0.125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-image-2":{"id":"openai/gpt-image-2","type":"image","name":"GPT Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"openai/tts-1-hd":{"id":"openai/tts-1-hd","type":"speech","name":"TTS-1 HD","description":"Text-to-speech model optimized for audio quality","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/tts-1-hd"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2-Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-4o-mini-fast":{"id":"openai/gpt-4o-mini-fast","name":"GPT-4o mini (Fast)","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"cost":{"input":0.25,"output":1,"cache_read":0.125},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":2.5,"output":20,"cache_read":0.25},"provider":{"body":{"service_tier":"priority"}}},"flex":{"cost":{"input":0.625,"output":5,"cache_read":0.0625},"provider":{"body":{"service_tier":"flex"}}}}},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-6.1-sol-fast":{"id":"openai/gpt-6.1-sol-fast","name":"GPT-6.1 Sol (Fast)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.4,"cache_write":10,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.4,"cache_write":10}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3 Pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":100000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.13},"canonical_model_id":"openai/gpt-5-codex"},"openai/tts-1":{"id":"openai/tts-1","type":"speech","name":"TTS-1","description":"Text-to-speech model optimized for low latency","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"openai/tts-1"},"openai/gpt-5.6-sol-fast":{"id":"openai/gpt-5.6-sol-fast","name":"GPT 5.6 Sol (Fast)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":8,"output":40,"cache_read":0.8,"cache_write":10,"tiers":[{"input":16,"output":60,"cache_read":1.6,"cache_write":20,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":16,"output":60,"cache_read":1.6,"cache_write":20}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-luna-fast":{"id":"openai/gpt-6-luna-fast","name":"GPT-6 Luna (Fast)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.5,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":0.4,"output":1.5,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT 5.6 Sol","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-6-astra-fast":{"id":"openai/gpt-6-astra-fast","name":"GPT-6 Astra (Fast)","description":"Fast variant of GPT-6 Astra for low-latency assistance and high-volume workloads.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":20,"output":100,"cache_read":2,"cache_write":25,"tiers":[{"input":40,"output":150,"cache_read":4,"cache_write":50,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":40,"output":150,"cache_read":4,"cache_write":50}},"canonical_model_id":"openai/gpt-6-astra-fast"},"openai/gpt-5-mini-fast":{"id":"openai/gpt-5-mini-fast","name":"GPT-5 mini (Fast)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.45,"output":3.6,"cache_read":0.045},"canonical_model_id":"openai/gpt-5-mini"},"zai/glm-5v-turbo":{"id":"zai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"zai/glm-4.5v":{"id":"zai/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":66000,"output":16000},"cost":{"input":0.6,"output":1.8,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5v"},"zai/glm-4.6":{"id":"zai/glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":96000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.6"},"zai/glm-4.7-flash":{"id":"zai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131000},"cost":{"input":0.07,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai/glm-4.5-air":{"id":"zai/glm-4.5-air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.2,"output":1.1,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai/glm-5.2":{"id":"zai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.8,"output":2.55,"cache_read":0.16},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-4.7":{"id":"zai/glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":120000},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"zai/glm-5.1":{"id":"zai/glm-5.1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":64000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai/glm-4.5":{"id":"zai/glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.5"},"zai/glm-5.2-fast":{"id":"zai/glm-5.2-fast","name":"GLM 5.2 Fast","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":2.8,"output":8.8,"cache_read":0.56},"canonical_model_id":"zhipuai/glm-5.2"},"zai/glm-5.3-flashx":{"id":"zai/glm-5.3-flashx","name":"GLM 5.3 FlashX","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.37,"output":1.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai/glm-5":{"id":"zai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202800,"output":131100},"cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"zai/glm-4.7-flashx":{"id":"zai/glm-4.7-flashx","name":"GLM 4.7 FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"zai/glm-5.3":{"id":"zai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.2,"output":4,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5.3-fast":{"id":"zai/glm-5.3-fast","name":"GLM 5.3 Fast","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.3"},"zai/glm-5-turbo":{"id":"zai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"zai/glm-5.3-flash":{"id":"zai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.05},"canonical_model_id":"inference-net/schematron-v2-small"},"microsoft/mai-image-2.6-flash":{"id":"microsoft/mai-image-2.6-flash","type":"image","name":"MAI-Image-2.6-Flash","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":1.75,"output":19}},"microsoft/mai-image-2.6":{"id":"microsoft/mai-image-2.6","type":"image","name":"MAI-Image-2.6","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"cost":{"input":5,"output":38}},"microsoft/mai-voice-2.1-flash":{"id":"microsoft/mai-voice-2.1-flash","type":"speech","name":"MAI-Voice-2.1-Flash","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-voice-2-flash":{"id":"microsoft/mai-voice-2-flash","type":"speech","name":"MAI-Voice-2-Flash","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-transcribe-2-streaming":{"id":"microsoft/mai-transcribe-2-streaming","type":"transcription","name":"MAI-Transcribe-2-Streaming","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mai","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-transcribe-2":{"id":"microsoft/mai-transcribe-2","type":"transcription","name":"MAI-Transcribe 2","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mai","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-03","last_updated":"2026-09-03","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-voice-2":{"id":"microsoft/mai-voice-2","type":"speech","name":"MAI-Voice-2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"microsoft/mai-voice-2.1":{"id":"microsoft/mai-voice-2.1","type":"speech","name":"MAI-Voice-2.1","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mai","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"poolside/laguna-s-2.1-free":{"id":"poolside/laguna-s-2.1-free","name":"Laguna S 2.1 Free","description":"Free provider route for experiments, demos, and cost-sensitive chat workloads","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.09,"output":0.18,"cache_read":0.009},"canonical_model_id":"poolside/laguna-s-2.1"},"interfaze/interfaze-beta":{"id":"interfaze/interfaze-beta","name":"Interfaze Beta","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"temperature":true,"release_date":"2025-10-07","last_updated":"2026-04-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":1.5,"output":3.5}},"anthropic/claude-opus-4.8-fast":{"id":"anthropic/claude-opus-4.8-fast","name":"Claude Opus 4.8 (Fast)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-opus-5-fast":{"id":"anthropic/claude-opus-5-fast","name":"Claude Opus 5 (Fast)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200001}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100001}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-5.5-fast":{"id":"anthropic/claude-opus-5.5-fast","name":"Claude Opus 5.5 (Fast)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":8,"output":40,"cache_read":0.4,"cache_write":10},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra-v2"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/namazu":{"id":"sakana/namazu","name":"Sakana Namazu","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272001}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 High Speed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131100},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-h3":{"id":"minimax/minimax-h3","type":"video","name":"MiniMax H3","description":"MiniMax general-purpose video model with multimodal references and synchronized stereo audio. Open H3-Base checkpoints; Context-IR and 2K regeneration are not included in the initial weights release.","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-31","last_updated":"2026-10-04","modalities":{"input":["text","image"],"output":["video"]},"open_weights":true,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"minimax/MiniMax-H3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"Minimax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.1-lightning":{"id":"minimax/minimax-m2.1-lightning","name":"MiniMax M2.1 Lightning","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-12-23","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":2.4,"cache_read":0.03,"cache_write":0.375}},"minimax/minimax-m2.5-highspeed":{"id":"minimax/minimax-m2.5-highspeed","name":"MiniMax M2.5 High Speed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131000},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131000},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-h3-max":{"id":"minimax/minimax-h3-max","type":"video","name":"MiniMax H3 Max","description":"Fast video model post-trained by fal.ai on MiniMax H3, with synchronized audio at 480p and 768p","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"minimax/MiniMax-H3-Max"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":196608},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"mistral/ministral-8b":{"id":"mistral/ministral-8b","name":"Ministral 8B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-16","last_updated":"2024-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4000},"cost":{"input":0.15,"output":0.15,"cache_read":0.015}},"mistral/ministral-14b":{"id":"mistral/ministral-14b","name":"Ministral 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-14b"},"mistral/codestral":{"id":"mistral/codestral","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-latest"},"mistral/mistral-small":{"id":"mistral/mistral-small","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"mistral/mistral-small-latest"},"mistral/voxtral-mini-tts-3":{"id":"mistral/voxtral-mini-tts-3","type":"speech","name":"Voxtral TTS 3","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"mistral/codestral-embed":{"id":"mistral/codestral-embed","type":"embedding","name":"Codestral Embed","description":"Embedding model for code search, retrieval, and code-aware RAG pipelines","family":"codestral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"mistral/codestral-embed-2505"},"mistral/mistral-large-3":{"id":"mistral/mistral-large-3","name":"Mistral Large 3","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"mistral/mistral-large-2512"},"mistral/mistral-large-4":{"id":"mistral/mistral-large-4","name":"Mistral Large 4","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"mistral/mistral-embed":{"id":"mistral/mistral-embed","type":"embedding","name":"Mistral Embed","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":3072},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"mistral/mistral-embed"},"mistral/voxtral-small-transcribe-3":{"id":"mistral/voxtral-small-transcribe-3","type":"transcription","name":"Voxtral Small Transcribe 3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"mistral/mistral-nemo":{"id":"mistral/mistral-nemo","name":"Mistral Nemo","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.029,"output":0.03},"canonical_model_id":"mistral/mistral-nemo"},"mistral/mistral-medium-3.5":{"id":"mistral/mistral-medium-3.5","name":"Mistral Medium Latest","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":256000},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-3-5"},"mistral/voxtral-mini-transcribe-3":{"id":"mistral/voxtral-mini-transcribe-3","type":"transcription","name":"Voxtral Mini Transcribe 3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"mistral/ministral-3b":{"id":"mistral/ministral-3b","name":"Ministral 3B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-16","last_updated":"2024-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4000},"cost":{"input":0.1,"output":0.1,"cache_read":0.01}},"callstack/apex":{"id":"callstack/apex","name":"Apex","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"release_date":"2026-10-08","last_updated":"2026-10-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.2}},"heygen/heygen-video-1":{"id":"heygen/heygen-video-1","type":"video","name":"HeyGen Video","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-09","last_updated":"2026-10-09","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"prodia/flux-fast-schnell":{"id":"prodia/flux-fast-schnell","type":"image","name":"Flux Schnell","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-02","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"scaledown/scaledown-summarize-1":{"id":"scaledown/scaledown-summarize-1","name":"Summarize-1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000},"cost":{"input":0.05,"output":0}},"scaledown/scaledown-extract-1":{"id":"scaledown/scaledown-extract-1","name":"Extract-1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":0.05,"output":0}},"bytedance/seedance-2.0-fast":{"id":"bytedance/seedance-2.0-fast","type":"video","name":"Seedance 2.0 Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-14","last_updated":"2026-04-14","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-v1.0-pro":{"id":"bytedance/seedance-v1.0-pro","type":"video","name":"Seedance v1.0 Pro","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-11","last_updated":"2025-06-11","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seed-1.6":{"id":"bytedance/seed-1.6","name":"Seed 1.6","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.25,"output":2,"cache_read":0.05,"tiers":[{"input":0.5,"output":4,"cache_read":0.05,"tier":{"type":"context","size":128001}}]},"canonical_model_id":"bytedance-seed/seed-1-6"},"bytedance/seedream-5.0-pro":{"id":"bytedance/seedream-5.0-pro","type":"image","name":"Seedream 5.0 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-11","last_updated":"2026-07-11","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedream-5.0-lite":{"id":"bytedance/seedream-5.0-lite","type":"image","name":"Seedream 5.0 Lite","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-13","last_updated":"2026-01-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedream-4.5":{"id":"bytedance/seedream-4.5","type":"image","name":"Seedream 4.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-11-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seed-1.8":{"id":"bytedance/seed-1.8","name":"Seed 1.8","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32768},"cost":{"input":0.25,"output":2,"cache_read":0.05,"tiers":[{"input":0.5,"output":4,"cache_read":0.05,"tier":{"type":"context","size":128001}}]},"canonical_model_id":"bytedance-seed/seed-1-8"},"bytedance/seedance-v1.0-pro-fast":{"id":"bytedance/seedance-v1.0-pro-fast","type":"video","name":"Seedance v1.0 Pro Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-24","last_updated":"2025-10-31","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-2.0":{"id":"bytedance/seedance-2.0","type":"video","name":"Seedance 2.0","description":"ByteDance unified multimodal audio-video model supporting text, image, video and audio references, editing, extension and synchronized stereo sound.","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-12","last_updated":"2026-10-04","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"bytedance-seed/seedance-2.0"},"bytedance/seedream-4.0":{"id":"bytedance/seedream-4.0","type":"image","name":"Seedream 4.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-09","last_updated":"2025-08-28","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-v1.5-pro":{"id":"bytedance/seedance-v1.5-pro","type":"video","name":"Seedance v1.5 Pro","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seedance-2.5":{"id":"bytedance/seedance-2.5","type":"video","name":"Seedance 2.5","description":"ByteDance multimodal audio-video model with up to 30-second single-pass generation, multi-round extension, reference inputs and timestamp-level editing.","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-31","last_updated":"2026-10-04","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"bytedance-seed/seedance-2.5"},"bytedance/seedance-2.0-mini":{"id":"bytedance/seedance-2.0-mini","type":"video","name":"Seedance 2.0 Mini","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-06-22","last_updated":"2026-06-22","modalities":{"input":["text","image","video","audio"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bytedance/seed-2.1-turbo":{"id":"bytedance/seed-2.1-turbo","name":"Seed 2.1 Turbo","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":2.5,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"cohere/rerank-v4-fast":{"id":"cohere/rerank-v4-fast","type":"reranking","name":"Cohere Rerank 4 Fast","description":"Low-latency multilingual reranker for high-throughput retrieval","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/rerank-v4.0-fast"},"cohere/embed-v4.0":{"id":"cohere/embed-v4.0","type":"embedding","name":"Embed v4.0","description":"Multimodal embedding model for text, images, and mixed text-and-image documents","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":1536},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/embed-v4.0"},"cohere/embed-v5.0-fast":{"id":"cohere/embed-v5.0-fast","type":"embedding","name":"Embed 5 Fast","description":"Faster, lighter Embed 5 model sharing Embed 5 Pro's embedding space","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":2048},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/embed-v5.0-fast"},"cohere/rerank-v5-pro":{"id":"cohere/rerank-v5-pro","type":"reranking","name":"Cohere Rerank 5 Pro","description":"Reranking model for improving retrieval quality in search and recommendation systems","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"cohere/rerank-v4-pro":{"id":"cohere/rerank-v4-pro","type":"reranking","name":"Cohere Rerank 4 Pro","description":"Multilingual reranker for maximum relevance on documents and semi-structured data","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/rerank-v4.0-pro"},"cohere/embed-v5.0-pro":{"id":"cohere/embed-v5.0-pro","type":"embedding","name":"Embed 5 Pro","description":"Cohere's most capable multimodal embedding model for text, images, and mixed documents","family":"cohere-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-30","last_updated":"2026-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":2048},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/embed-v5.0-pro"},"cohere/rerank-v5-fast":{"id":"cohere/rerank-v5-fast","type":"reranking","name":"Cohere Rerank 5 Fast","description":"Reranking model for improving retrieval quality in search and recommendation systems","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a"},"cohere/rerank-v3.5":{"id":"cohere/rerank-v3.5","type":"reranking","name":"Cohere Rerank 3.5","description":"Multilingual reranker for documents and semi-structured data","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-12-02","last_updated":"2024-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":4096,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"},"canonical_model_id":"cohere/rerank-v3.5"},"bfl/flux-kontext-max":{"id":"bfl/flux-kontext-max","type":"image","name":"FLUX.1 Kontext Max","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-29","last_updated":"2025-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-pro-1.1-ultra":{"id":"bfl/flux-pro-1.1-ultra","type":"image","name":"FLUX1.1 [pro] Ultra","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-11-01","last_updated":"2024-11","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-3-video":{"id":"bfl/flux-3-video","type":"video","name":"Flux 3","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-04","last_updated":"2026-08-04","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-pro-1.1":{"id":"bfl/flux-pro-1.1","type":"image","name":"FLUX1.1 [pro]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-02","last_updated":"2024-10","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-klein-4b":{"id":"bfl/flux-2-klein-4b","type":"image","name":"FLUX.2 [klein] 4B","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-3-image-klein":{"id":"bfl/flux-3-image-klein","type":"image","name":"FLUX 3 Image Klein","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-max":{"id":"bfl/flux-2-max","type":"image","name":"FLUX.2 [max]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":67300,"output":67300},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-flex":{"id":"bfl/flux-2-flex","type":"image","name":"FLUX.2 [flex]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-pro":{"id":"bfl/flux-2-pro","type":"image","name":"FLUX.2 [pro]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":67300,"output":67300},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-3-image":{"id":"bfl/flux-3-image","type":"image","name":"FLUX 3 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-pro-1.0-fill":{"id":"bfl/flux-pro-1.0-fill","type":"image","name":"FLUX.1 Fill [pro]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-01","last_updated":"2024-10","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-2-klein-9b":{"id":"bfl/flux-2-klein-9b","type":"image","name":"FLUX.2 [klein] 9B","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"bfl/flux-kontext-pro":{"id":"bfl/flux-kontext-pro","type":"image","name":"FLUX.1 Kontext Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-29","last_updated":"2025-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":512,"output":0},"provider":{"api":"https://ai-gateway.vercel.sh/v4/ai"}},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"Morph v3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":0.8,"output":1.2}},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"Morph v3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.9,"output":1.9}}}},"anthropic":{"id":"anthropic","env":["ANTHROPIC_API_KEY"],"npm":"@ai-sdk/anthropic","name":"Anthropic","doc":"https://docs.anthropic.com/en/docs/about-claude/models","models":{"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-29","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-07","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-04","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75}},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":8,"output":40,"cache_read":0.4,"cache_write":10},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25}},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-14","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25}}}},"vispark":{"id":"vispark","env":["VISPARK_LAB_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lab.vispark.in/v1","name":"Vispark","doc":"https://lab.vispark.in/#vision","models":{"vispark/vision-small":{"id":"vispark/vision-small","name":"Vision Small","description":"Fast, low-cost multimodal model for understanding text, images, audio, video, and PDFs, with tool calling and a 1M-token context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.05,"output":3.16},"canonical_model_id":"vispark/vision-small"},"vispark/vision-large":{"id":"vispark/vision-large","name":"Vision Large","description":"Most capable Vision model for complex reasoning, detailed media analysis, and structured output over a 1M-token context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":7.37,"output":22.11},"canonical_model_id":"vispark/vision-large"},"vispark/vision-medium":{"id":"vispark/vision-medium","name":"Vision Medium","description":"Balanced multimodal model pairing a 1M-token context window with deeper reasoning for analysis, content creation, and tool use across text, image, audio, video, and PDF inputs.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":4.21,"output":12.63},"canonical_model_id":"vispark/vision-medium"}}},"sensenova":{"id":"sensenova","env":["SENSENOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token.sensenova.cn/v1","name":"SenseNova (China)","doc":"https://platform.sensenova.cn/docs","models":{"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash is a high-performance conversational model by DeepSeek, supporting thinking/non-thinking modes, 1M context window, and tool calling.","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"moonshotai/kimi-k3"},"sensenova-6.8-flash-lite":{"id":"sensenova-6.8-flash-lite","name":"SenseNova 6.8 Flash Lite","description":"SenseNova lightweight multimodal agent model for real-world complex tasks, data analysis, and complex information presentation","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-11","last_updated":"2026-08-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0,"cache_read":0}},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"deepseek/deepseek-v4-flash"}}},"alibaba-token-plan":{"id":"alibaba-token-plan","env":["ALIBABA_TOKEN_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1","name":"Alibaba Token Plan","doc":"https://www.alibabacloud.com/help/en/model-studio/token-plan-overview","models":{"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"happyhorse-1.1-r2v":{"id":"happyhorse-1.1-r2v","type":"video","name":"HappyHorse 1.1 Reference-to-Video","description":"Video model for reference-guided video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"qwen3.8-max-preview":{"id":"qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"qwen-image-2.0-pro":{"id":"qwen-image-2.0-pro","type":"image","name":"Qwen Image 2.0 Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen-image-2.0":{"id":"qwen-image-2.0","type":"image","name":"Qwen Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"happyhorse-1.1-i2v":{"id":"happyhorse-1.1-i2v","type":"video","name":"HappyHorse 1.1 Image-to-Video","description":"Video model for image-to-video generation","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["image","text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-plus"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"wan2.7-image-pro":{"id":"wan2.7-image-pro","type":"image","name":"Wan2.7 Image Pro","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":131072}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.6-flash"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5"},"happyhorse-1.1-t2v":{"id":"happyhorse-1.1-t2v","type":"video","name":"HappyHorse 1.1 Text-to-Video","description":"Video model for prompt-driven text-to-video generation","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-17","last_updated":"2026-07-17","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"provider":{"npm":"@ai-sdk/alibaba","api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com"},"cost":{"input":0,"output":0}},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-max"},"wan2.7-image":{"id":"wan2.7-image","type":"image","name":"Wan2.7 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":0},"provider":{"api":"https://token-plan.ap-southeast-1.maas.aliyuncs.com/api/v1"},"cost":{"input":0,"output":0}},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-03","last_updated":"2025-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"status":"deprecated","cost":{"input":0,"output":0}},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","min":0,"max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.8-max"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"input":196601,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"302ai":{"id":"302ai","env":["302AI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.302.ai/v1","name":"302.AI","doc":"https://doc.302.ai","models":{"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.72,"output":3.2,"tiers":[{"input":1.1,"output":3.8,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"claude-opus-4-1-20250805-thinking":{"id":"claude-opus-4-1-20250805-thinking","name":"claude-opus-4-1-20250805-thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"claude-sonnet-4-5-20250929-thinking":{"id":"claude-sonnet-4-5-20250929-thinking","name":"claude-sonnet-4-5-20250929-thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.29,"output":0.86,"tiers":[{"input":0.58,"output":1.72,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5v"},"glm-4.6":{"id":"glm-4.6","name":"glm-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.286,"output":1.142,"tiers":[{"input":0.572,"output":2.29,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"gpt-4.1-mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6},"canonical_model_id":"openai/gpt-4.1-mini"},"doubao-seed-1-6-thinking-250715":{"id":"doubao-seed-1-6-thinking-250715","name":"doubao-seed-1-6-thinking-250715","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16000},"cost":{"input":0.121,"output":1.21}},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"gpt-4.1-nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"openai/gpt-4.1-nano"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":4,"output":10,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-4-argon"},"gpt-5.1":{"id":"gpt-5.1","name":"gpt-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.228,"output":0.92,"cache_read":0.023,"cache_write":0.286,"tiers":[{"input":0.68,"output":2.74,"cache_read":0.068,"cache_write":0.86,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.68,"output":2.74,"cache_read":0.068,"cache_write":0.86}},"canonical_model_id":"alibaba/qwen3.7-plus"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"gemini-3-flash-preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3-flash-preview"},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"ministral-14b-2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.33,"output":0.33},"canonical_model_id":"mistral/ministral-14b-2512"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"gpt-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":0,"tiers":[{"input":5,"output":22.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5}},"canonical_model_id":"openai/gpt-5.4"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"claude-opus-4-5-20251101","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1},"canonical_model_id":"openai/gpt-6-astra"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"gpt-5.6-luna-pro":{"id":"gpt-5.6-luna-pro","name":"gpt-5.6-luna-pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen3-max-2025-09-23":{"id":"qwen3-max-2025-09-23","name":"qwen3-max-2025-09-23","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":258048,"output":65536},"cost":{"input":0.86,"output":3.43,"tiers":[{"input":1.43,"output":5.72,"tier":{"type":"context","size":32000}},{"input":2.15,"output":8.58,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"glm-4.7":{"id":"glm-4.7","name":"glm-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.286,"output":1.142,"tiers":[{"input":0.572,"output":2.29,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"tiers":[{"input":4,"output":18,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gpt-4.1":{"id":"gpt-4.1","name":"gpt-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8},"canonical_model_id":"openai/gpt-4.1"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"gemini-2.5-flash-preview-09-2025":{"id":"gemini-2.5-flash-preview-09-2025","name":"gemini-2.5-flash-preview-09-2025","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.33,"output":1.32},"canonical_model_id":"minimax/MiniMax-M2"},"gemini-3.5-flash-thinking":{"id":"gemini-3.5-flash-thinking","name":"gemini-3.5-flash-thinking","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9},"canonical_model_id":"google/gemini-3.5-flash"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"grok-4-1-fast-reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"grok-4.1":{"id":"grok-4.1","name":"grok-4.1","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":2,"output":10}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"cost":{"input":0.72,"output":2.88,"cache_read":0.144,"cache_write":0.9,"tiers":[{"input":1.44,"output":5.76,"cache_read":0.144,"cache_write":0.9,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":1.44,"output":5.76,"cache_read":0.144,"cache_write":0.9}},"canonical_model_id":"minimax/MiniMax-M3"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"claude-opus-4-1-20250805","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gpt-5-mini":{"id":"gpt-5-mini","name":"gpt-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"gpt-5-thinking":{"id":"gpt-5-thinking","name":"gpt-5-thinking","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":0.5},"canonical_model_id":"xai/grok-4.7"},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"grok-4-fast-non-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.627,"output":3.3},"canonical_model_id":"moonshotai/kimi-k2.5"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.5"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"gemini-2.5-flash-image","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30},"canonical_model_id":"google/gemini-2.5-flash-image"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"claude-sonnet-4-6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"gemini-2.5-flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"claude-haiku-4-5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-opus-4-7-thinking":{"id":"claude-opus-4-7-thinking","name":"claude-opus-4-7-thinking","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-7"},"kimi-k2-0905-preview":{"id":"kimi-k2-0905-preview","name":"kimi-k2-0905-preview","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.62854,"output":2.5146},"canonical_model_id":"moonshotai/kimi-k2-0905"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02}},"canonical_model_id":"openai/gpt-6-luna"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.286,"output":1.143,"tiers":[{"input":0.5714,"output":2.29,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"gpt-5.2-chat-latest":{"id":"gpt-5.2-chat-latest","name":"gpt-5.2-chat-latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"doubao-seed-1-8-251215":{"id":"doubao-seed-1-8-251215","name":"doubao-seed-1-8-251215","description":"ByteDance Seed model for multimodal reasoning, long-context analysis, and agent workflows","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":224000,"output":64000},"cost":{"input":0.1143,"output":0.286,"tiers":[{"input":0.1715,"output":2.286,"tier":{"type":"context","size":32000}},{"input":0.343,"output":3.43,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-8"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-5-thinking":{"id":"claude-opus-5-thinking","name":"claude-opus-5-thinking","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5},"canonical_model_id":"xai/grok-4.3"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"gpt-5.4-mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":1.8,"tiers":[{"input":1.2,"output":7.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":7.2}},"canonical_model_id":"alibaba/qwen3.6-plus"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"gemini-2.5-pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":1.25,"output":10,"tiers":[{"input":2.5,"output":15,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15}},"canonical_model_id":"google/gemini-2.5-pro"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"qwen3-235b-a22b-instruct-2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":65536},"cost":{"input":0.29,"output":1.143},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-5.2":{"id":"gpt-5.2","name":"gpt-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2"},"gpt-5.6-terra-pro":{"id":"gpt-5.6-terra-pro","name":"gpt-5.6-terra-pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-terra"},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.12,"output":0.69,"tiers":[{"input":0.29,"output":1.72,"tier":{"type":"context","size":128000}},{"input":0.58,"output":3.44,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.5-plus"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"claude-sonnet-4-5-20250929","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-sonnet-4-6-thinking":{"id":"claude-sonnet-4-6-thinking","name":"claude-sonnet-4-6-thinking","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5-pro":{"id":"gpt-5-pro","name":"gpt-5-pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000},"cost":{"input":15,"output":120,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-pro"},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.19,"output":1.13,"cache_read":0.0171,"cache_write":0.214,"tiers":[{"input":0.76,"output":4.52,"cache_read":0.0686,"cache_write":0.857,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.76,"output":4.52,"cache_read":0.0686,"cache_write":0.857}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen3-coder-480b-a35b-instruct":{"id":"qwen3-coder-480b-a35b-instruct","name":"qwen3-coder-480b-a35b-instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.86,"output":3.43,"tiers":[{"input":1.29,"output":5.15,"tier":{"type":"context","size":32000}},{"input":2.15,"output":8.58,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"glm-5":{"id":"glm-5","name":"glm-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.6,"tiers":[{"input":0.9,"output":3.2,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"MiniMax-M1":{"id":"MiniMax-M1","name":"MiniMax-M1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-16","last_updated":"2025-06-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.132,"output":1.254}},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.8,"output":5.3},"canonical_model_id":"alibaba/qwen3.7-max"},"gemini-2.5-flash-nothink":{"id":"gemini-2.5-flash-nothink","name":"gemini-2.5-flash-nothink","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.145,"output":0.43,"tiers":[{"input":0.29,"output":0.86,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"gpt-5.1-chat-latest":{"id":"gpt-5.1-chat-latest","name":"gpt-5.1-chat-latest","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"gpt-5.4-nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-4o":{"id":"gpt-4o","name":"gpt-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-4o"},"grok-4.20-beta-0309-reasoning":{"id":"grok-4.20-beta-0309-reasoning","name":"grok-4.20-beta-0309-reasoning","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"qwen3-235b-a22b":{"id":"qwen3-235b-a22b","name":"Qwen3-235B-A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.29,"output":2.86},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"claude-haiku-4-5-20251001","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"deepseek-flash":{"id":"deepseek-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"gemini-2.5-flash-lite-preview-09-2025":{"id":"gemini-2.5-flash-lite-preview-09-2025","name":"gemini-2.5-flash-lite-preview-09-2025","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1,"output":0.4}},"gemini-2.0-flash-lite":{"id":"gemini-2.0-flash-lite","name":"gemini-2.0-flash-lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-11","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":8192},"cost":{"input":0.075,"output":0.3},"canonical_model_id":"google/gemini-2.0-flash-lite"},"deepseek-v3.2-thinking":{"id":"deepseek-v3.2-thinking","name":"DeepSeek-V3.2-Thinking","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.29,"output":0.43},"canonical_model_id":"deepseek/deepseek-v3.2"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8},"canonical_model_id":"openai/o3"},"grok-4-fast-reasoning":{"id":"grok-4-fast-reasoning","name":"grok-4-fast-reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4-fast-reasoning"},"gpt-5.6-sol-pro":{"id":"gpt-5.6-sol-pro","name":"gpt-5.6-sol-pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.6-sol"},"glm-5-turbo":{"id":"glm-5-turbo","name":"glm-5-turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.72,"output":3.2,"tiers":[{"input":1.1,"output":3.8,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5-turbo"},"doubao-seed-1-6-vision-250815":{"id":"doubao-seed-1-6-vision-250815","name":"doubao-seed-1-6-vision-250815","description":"ByteDance Seed multimodal model for image understanding, visual reasoning, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-08-15","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.114,"output":1.143},"canonical_model_id":"bytedance-seed/seed-1-6-vision"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"kimi-k2-thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.575,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"gemini-3-pro-image-preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":64000},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image-preview"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.28,"output":1.7},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"deepseek-v3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.29,"output":0.43},"canonical_model_id":"deepseek/deepseek-v3.2"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.16,"output":6.36,"cache_read":0.168,"cache_write":2.58},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"claude-opus-4-7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-7"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.075,"output":0.25,"cache_read":0.015},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gpt-5":{"id":"gpt-5","name":"gpt-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"gemini-3-pro-preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":12,"tiers":[{"input":4,"output":18,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18}},"canonical_model_id":"google/gemini-3-pro-preview"},"qwen3.7-max-2026-06-08":{"id":"qwen3.7-max-2026-06-08","name":"qwen3.7-max-2026-06-08","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":1.8,"output":5.3},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3-30B-A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0.11,"output":1.08},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"mistral-large-2512":{"id":"mistral-large-2512","name":"mistral-large-2512","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"cost":{"input":1.1,"output":3.3},"canonical_model_id":"mistral/mistral-large-2512"},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.06,"output":0.46,"tiers":[{"input":0.23,"output":1.83,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"gpt-5.3-chat-latest":{"id":"gpt-5.3-chat-latest","name":"GPT-5.3 Chat (latest)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.5"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.1-flash-image-preview":{"id":"gemini-3.1-flash-image-preview","type":"image","name":"gemini-3.1-flash-image-preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":60},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"}}},"sakana":{"id":"sakana","env":["SAKANA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.sakana.ai/v1","name":"Sakana AI","doc":"https://console.sakana.ai/models","models":{"fugu":{"id":"fugu","name":"Fugu","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"provider":{"shape":"responses"},"canonical_model_id":"sakana/fugu"},"fugu-ultra-20260615":{"id":"fugu-ultra-20260615","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"provider":{"shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"sakana-namazu":{"id":"sakana-namazu","name":"Sakana Namazu","description":"Japanese-specialized reasoning model based on Kimi K2.6 and tuned for Japanese language, culture, and business workflows","family":"sakana-namazu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.15},"canonical_model_id":"sakana/sakana-namazu"},"fugu-ultra":{"id":"fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"provider":{"shape":"responses"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"}}},"wallaby":{"id":"wallaby","env":["WALLABY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.wallabytoken.com/v1","name":"Wallaby","doc":"https://wallabytoken.com/docs","models":{"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.7,"output":13.5,"cache_read":0.27},"canonical_model_id":"moonshotai/kimi-k3"}}},"minimax":{"id":"minimax","env":["MINIMAX_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.minimax.io/anthropic/v1","name":"MiniMax (minimax.io)","doc":"https://platform.minimax.io/docs/guides/quickstart","models":{"MiniMax-M2.7-highspeed":{"id":"MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375}},"MiniMax-M3.1-Flash":{"id":"MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375}},"MiniMax-M2":{"id":"MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2}},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-25","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}}},"MiniMax-M2.5-highspeed":{"id":"MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375}},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375}}}},"mistral":{"id":"mistral","env":["MISTRAL_API_KEY"],"npm":"@ai-sdk/mistral","name":"Mistral","doc":"https://docs.mistral.ai/getting-started/models/","models":{"voxtral-small-2507":{"id":"voxtral-small-2507","name":"Voxtral Small","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.1,"output":0.4}},"mistral-small-2506":{"id":"mistral-small-2506","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.1,"output":0.3}},"zai-glm-5-2":{"id":"zai-glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"devstral-latest":{"id":"devstral-latest","name":"Devstral 2","description":"Legacy model retained for compatibility with older integrations","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"Ministral 3 14B","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.2,"cache_read":0.02},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"devstral-medium-latest":{"id":"devstral-medium-latest","name":"Devstral 2 (latest)","description":"Legacy model retained for compatibility with older integrations","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"labs-leanstral-1-5-1":{"id":"labs-leanstral-1-5-1","name":"Leanstral 1.5","description":"An updated Lean 4 formal proof engineering model optimised for automated theorem proving and autoformalization. 119B total parameters, 6.5B active.","family":"leanstral","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":0,"output":0}},"devstral-2512":{"id":"devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"magistral-medium-latest":{"id":"magistral-medium-latest","name":"Magistral Medium (latest)","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-03-17","last_updated":"2025-03-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":2,"output":5}},"labs-devstral-small-2512":{"id":"labs-devstral-small-2512","name":"Devstral Small 2","description":"Legacy model retained for compatibility with older integrations","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0}},"open-mistral-nemo":{"id":"open-mistral-nemo","name":"Open Mistral Nemo","description":"Legacy model retained for compatibility with older integrations","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.15,"output":0.15}},"glm-5-2":{"id":"glm-5-2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.2"},"ministral-3b-latest":{"id":"ministral-3b-latest","name":"Ministral 3B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-01","last_updated":"2024-10-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":128000},"cost":{"input":0.1,"output":0.1,"cache_read":0.01}},"codestral-2508":{"id":"codestral-2508","name":"Codestral 25.08","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"cost":{"input":0.3,"output":0.9,"cache_read":0.03},"canonical_model_id":"mistral/codestral-2508"},"ministral-8b-latest":{"id":"ministral-8b-latest","name":"Ministral 8B (latest)","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-01","last_updated":"2024-10-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.15,"output":0.15,"cache_read":0.015}},"ministral-8b-2512":{"id":"ministral-8b-2512","name":"Ministral 3 8B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.15,"output":0.15,"cache_read":0.015},"canonical_model_id":"mistral/ministral-3-8b-instruct-2512"},"voxtral-mini-tts-latest":{"id":"voxtral-mini-tts-latest","type":"speech","name":"Voxtral Mini TTS (latest)","description":"Multilingual text-to-speech model with zero-shot voice cloning","family":"voxtral","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-23","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["audio"]},"open_weights":true,"limit":{"context":0,"output":0}},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"status":"beta","cost":{"input":1.36,"output":4.18,"cache_read":0.14},"canonical_model_id":"mistral/mistral-large-4"},"pixtral-12b":{"id":"pixtral-12b","name":"Pixtral 12B","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-09-01","last_updated":"2024-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.15,"output":0.15}},"zai-glm-5-3":{"id":"zai-glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.14},"canonical_model_id":"zhipuai/glm-5.3"},"mistral-embed":{"id":"mistral-embed","type":"embedding","name":"Mistral Embed","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":3072},"cost":{"input":0.1,"output":0}},"mistral-medium-2508":{"id":"mistral-medium-2508","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.4,"output":2}},"mistral-small-latest":{"id":"mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015}},"mistral-large-latest":{"id":"mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05}},"voxtral-small-latest":{"id":"voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32000},"cost":{"input":0.1,"output":0.4}},"voxtral-mini-latest":{"id":"voxtral-mini-latest","type":"transcription","name":"Voxtral Mini (latest)","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"voxtral","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"mistral-medium-latest":{"id":"mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"mistral/mistral-medium-latest"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015}},"codestral-latest":{"id":"codestral-latest","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"cost":{"input":0.3,"output":0.9,"cache_read":0.03}},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.5,"output":1.5,"cache_read":0.05}},"zai-glm-5-3-flash":{"id":"zai-glm-5-3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"mistral-medium-2604":{"id":"mistral-medium-2604","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.5,"output":7.5,"cache_read":0.15}},"mistral-medium-2505":{"id":"mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":2}},"ministral-3b-2512":{"id":"ministral-3b-2512","name":"Ministral 3 3B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":262144},"cost":{"input":0.1,"output":0.1,"cache_read":0.01},"canonical_model_id":"mistral/ministral-3-3b-instruct-2512"}}},"zhipuai":{"id":"zhipuai","env":["ZHIPU_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://open.bigmodel.cn/api/paas/v4","name":"Zhipu AI","doc":"https://docs.bigmodel.cn/cn/guide/start/pricing","models":{"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.74698,"output":3.28669,"cache_read":0.17927,"cache_write":0,"tiers":[{"input":1.04577,"output":3.88427,"cache_read":0.26891,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.29879,"output":0.89637,"cache_read":0.05976,"cache_write":0,"tiers":[{"input":0.59758,"output":1.79274,"cache_read":0.11952,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5v"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"glm-4.5-air":{"id":"glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.11952,"output":0.89637,"cache_read":0.0239,"cache_write":0,"tiers":[{"input":0.17927,"output":1.19516,"cache_read":0.03585,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5-air"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.19516,"output":4.18306,"cache_read":0.29879,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"glm-4.5-flash":{"id":"glm-4.5-flash","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.44819,"output":2.09153,"cache_read":0.08964,"cache_write":0,"tiers":[{"input":0.59758,"output":2.39032,"cache_read":0.11952,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.89637,"output":3.58548,"cache_read":0.19421,"cache_write":0,"tiers":[{"input":1.19516,"output":4.18306,"cache_read":0.29879,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"glm-4.5":{"id":"glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"cost":{"input":0.6,"output":2.2,"cache_read":0.11,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.5"},"glm-5.3-flashx":{"id":"glm-5.3-flashx","name":"GLM-5.3-FlashX","description":"High-speed GLM-5.3-Flash serving option for coding and agent workflows","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.29879,"output":1.04577,"cache_read":0.08516,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"glm-4.6v-flash":{"id":"glm-4.6v-flash","name":"GLM-4.6V-Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"glm-5":{"id":"glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.59758,"output":2.68911,"cache_read":0.1494,"cache_write":0,"tiers":[{"input":0.89637,"output":3.28669,"cache_read":0.22409,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"glm-4.7-flashx":{"id":"glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.0747,"output":0.44819,"cache_read":0.01494,"cache_write":0},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.19516,"output":4.18306,"cache_read":0.29879,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.1494,"output":0.44819,"cache_read":0.02988,"cache_write":0,"tiers":[{"input":0.29879,"output":0.89637,"cache_read":0.05976,"cache_write":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11952,"output":0.41831,"cache_read":0.03436,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"fastrouter":{"id":"fastrouter","env":["FASTROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://go.fastrouter.ai/api/v1","name":"FastRouter","doc":"https://fastrouter.ai/models","models":{"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.05,"output":3.5,"cache_read":0.205},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.08,"cache_read":0.12},"canonical_model_id":"zhipuai/glm-5"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/veo3.1-fast":{"id":"google/veo3.1-fast","type":"video","name":"Veo 3.1 Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-01-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"google/veo-3.1-fast-generate-preview"},"google/veo3.1-lite":{"id":"google/veo3.1-lite","type":"video","name":"Veo 3.1 Lite","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"google/veo-3.1-lite-generate-preview"},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/veo3.1":{"id":"google/veo3.1","type":"video","name":"Veo 3.1","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"google/veo-3.1-generate-preview"},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.145},"canonical_model_id":"deepseek/deepseek-v4-pro"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.75,"output":3.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"sarvam/sarvam-105b":{"id":"sarvam/sarvam-105b","name":"Sarvam 105B","description":"Flagship Indian-language reasoning model for enterprise multilingual applications","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.04,"output":0.16,"cache_read":0.025},"canonical_model_id":"sarvam/sarvam-105b"},"leonardo-ai/lucid-realism":{"id":"leonardo-ai/lucid-realism","name":"Lucid Realism","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"lucid","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-01","last_updated":"2025-06-01","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":4096,"output":0}},"leonardo-ai/lucid-origin":{"id":"leonardo-ai/lucid-origin","name":"Lucid Origin","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"lucid","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-01","last_updated":"2025-06-01","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":4096,"output":0}},"wanx/wan-v2-6":{"id":"wanx/wan-v2-6","type":"video","name":"Wan 2.6","description":"Video model for text- and image-to-video with audio input and synchronized sound","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-03","last_updated":"2025-12-03","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":400000,"output":0},"canonical_model_id":"alibaba/wan2.6"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.09,"output":0.36},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.07,"output":0.25,"cache_read":0.007},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-10-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-10-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-image-2":{"id":"openai/gpt-image-2","type":"image","name":"GPT Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"openai/gpt-image-2"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-10-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4":{"id":"x-ai/grok-4","name":"Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.75,"cache_write":15}},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2,"tiers":[{"input":2,"output":4,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2,"output":4,"cache_read":0.4}},"canonical_model_id":"xai/grok-build-0.1"},"bytedance/seedance-2":{"id":"bytedance/seedance-2","type":"video","name":"Seedance 2","description":"ByteDance unified multimodal audio-video model supporting text, image, video and audio references, editing, extension and synchronized stereo sound.","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-12","last_updated":"2026-10-04","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":4096,"output":0},"canonical_model_id":"bytedance-seed/seedance-2.0"}}},"ai-router":{"id":"ai-router","env":["AI_ROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.ai-router.dev/v1","name":"AI-ROUTER","doc":"https://ai-router.dev/openai-compatible-api-gateway/","models":{"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":1,"output":6,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"openai/gpt-5.6-sol"}}},"moark":{"id":"moark","env":["MOARK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://moark.com/v1","name":"Moark","doc":"https://moark.com/docs/openapi/v1#tag/%E6%96%87%E6%9C%AC%E7%94%9F%E6%88%90","models":{"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.98,"output":4},"canonical_model_id":"moonshotai/kimi-k2.6"},"MiniMax-M2.1":{"id":"MiniMax-M2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.29,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"DeepSeek-V4-Pro":{"id":"DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.8,"output":3.6},"canonical_model_id":"deepseek/deepseek-v4-pro"},"DeepSeek-V4-Flash":{"id":"DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.5,"output":0.5},"canonical_model_id":"deepseek/deepseek-v4-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.15,"output":0.3},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"GLM-5.3":{"id":"GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.2,"output":4.18},"canonical_model_id":"zhipuai/glm-5.3"},"GLM-4.7":{"id":"GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.42,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"Qwen3.6-Flash":{"id":"Qwen3.6-Flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.18,"output":1.03},"canonical_model_id":"alibaba/qwen3.6-flash"},"MiniMax-M2.7":{"id":"MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.29,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"}}},"deepinfra":{"id":"deepinfra","env":["DEEPINFRA_API_KEY"],"npm":"@ai-sdk/deepinfra","name":"Deep Infra","doc":"https://deepinfra.com/models","models":{"mistralai/Mistral-Small-24B-Instruct-2501":{"id":"mistralai/Mistral-Small-24B-Instruct-2501","name":"Mistral Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.08,"output":0.16},"canonical_model_id":"google/gemma-3-27b-it"},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.15},"canonical_model_id":"google/gemma-3-12b-it"},"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-4b-it"},"google/gemma-4-E4B-it":{"id":"google/gemma-4-E4B-it","name":"Gemma 4 E4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.02,"output":0.1},"canonical_model_id":"google/gemma-4-E4B-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"nvidia/Llama-3.3-Nemotron-Super-49B-v1.5":{"id":"nvidia/Llama-3.3-Nemotron-Super-49B-v1.5","name":"Llama 3.3 Nemotron Super 49B v1.5","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":0.4},"canonical_model_id":"nvidia/llama-3.3-nemotron-super-49b-v1.5"},"nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning":{"id":"nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.2,"output":0.8},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"nvidia/Nemotron-3-Nano-30B-A3B":{"id":"nvidia/Nemotron-3-Nano-30B-A3B","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.05,"output":0.2,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":512000},"cost":{"input":0.28,"output":1.1,"cache_read":0.056},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.25,"output":1,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M2.7"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-06","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"status":"deprecated","cost":{"input":0.15,"output":1.15,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":1048576},"status":"deprecated","cost":{"input":0.95,"output":4.05,"cache_read":0.16},"canonical_model_id":"thinkingmachines/inkling"},"tencent/Hy4-preview":{"id":"tencent/Hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/Hy3":{"id":"tencent/Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":128000},"status":"deprecated","cost":{"input":0.13,"output":0.53,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"meta-llama/Llama-Guard-4-12B":{"id":"meta-llama/Llama-Guard-4-12B","name":"Llama Guard 4 12B","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.18,"output":0.18},"canonical_model_id":"meta/llama-guard-4-12b"},"meta-llama/Llama-3.3-70B-Instruct-Turbo":{"id":"meta-llama/Llama-3.3-70B-Instruct-Turbo","name":"Llama 3.3 70B Turbo","description":"Compact Llama instruction model for fast chat and local deployment","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.32},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/Llama-4-Scout-17B-16E-Instruct":{"id":"meta-llama/Llama-4-Scout-17B-16E-Instruct","name":"Llama 4 Scout 17B","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":327680,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"id":"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","name":"Llama 4 Maverick 17B FP8","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":0.8},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.75,"output":3.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.68,"output":3.4,"cache_read":0.136,"cache_write":0.85},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.85,"output":14.25,"cache_read":0.285,"cache_write":3.5625},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26,"output":2.6},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-VL-235B-A22B-Instruct":{"id":"Qwen/Qwen3-VL-235B-A22B-Instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":0.88,"cache_read":0.11},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.09,"output":0.55},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-30B-A3B":{"id":"Qwen/Qwen3-30B-A3B","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.12,"output":0.5},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":16384},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.32,"output":3.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.2,"output":2.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"status":"deprecated","cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":1.1},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo","name":"Qwen3 Coder 480B A35B Instruct Turbo","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":66536},"cost":{"input":0.3,"output":1,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen 3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.1,"output":0.95,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-Max":{"id":"Qwen/Qwen3-Max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"status":"deprecated","cost":{"input":1.2,"output":6,"cache_read":0.24,"tiers":[{"input":2.4,"output":12,"cache_read":0.48,"tier":{"type":"context","size":32000}},{"input":3,"output":15,"cache_read":0.6,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.29,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen 3.5 397B A17B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.45,"output":3,"cache_read":0.22},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"Qwen/Qwen3.8-Max":{"id":"Qwen/Qwen3.8-Max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":1.65,"output":4.951,"cache_read":0.206},"canonical_model_id":"alibaba/qwen3.8-max"},"Qwen/Qwen3.8-Flash":{"id":"Qwen/Qwen3.8-Flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.113,"output":0.382,"cache_read":0.0141},"canonical_model_id":"alibaba/qwen3.8-flash"},"Qwen/Qwen3.7-Max":{"id":"Qwen/Qwen3.7-Max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.5,"tiers":[{"input":5,"output":15,"cache_read":1,"tier":{"type":"context","size":32000}},{"input":6.25,"output":18.5,"cache_read":1.25,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.7-max"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":8192},"cost":{"input":0.25,"output":0.95,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"status":"deprecated","cost":{"input":0.24,"output":0.9,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-V4.1-Flash":{"id":"deepseek-ai/DeepSeek-V4.1-Flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"id":"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"cost":{"input":0.09,"output":0.18,"cache_read":0.018,"cache_write":0.1125},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":8192},"cost":{"input":0.32,"output":0.89},"canonical_model_id":"deepseek/deepseek-v3"},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":64000},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.06,"output":0.18,"cache_read":0.015},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4.1-Flash-1017":{"id":"deepseek-ai/DeepSeek-V4.1-Flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.2,"output":0.6,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek-R1-0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":64000},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.3,"output":2.6,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.037,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.03,"output":0.14},"canonical_model_id":"openai/gpt-oss-20b"},"stepfun-ai/Step-3.7-Flash":{"id":"stepfun-ai/Step-3.7-Flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"status":"deprecated","cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun-ai/Step-5-Preview":{"id":"stepfun-ai/Step-5-Preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.1},"canonical_model_id":"stepfun/step-5-preview"},"XiaomiMiMo/MiMo-V2.5-Pro":{"id":"XiaomiMiMo/MiMo-V2.5-Pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":16384},"status":"deprecated","cost":{"input":1,"output":3,"cache_read":0.2},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"XiaomiMiMo/MiMo-V2.6-Pro":{"id":"XiaomiMiMo/MiMo-V2.6-Pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"XiaomiMiMo/MiMo-V2.5":{"id":"XiaomiMiMo/MiMo-V2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"XiaomiMiMo/MiMo-V2.6-Flash":{"id":"XiaomiMiMo/MiMo-V2.6-Flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.9,"output":4,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0.6,"output":2.08,"cache_read":0.12},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":1.05,"output":3.5,"cache_read":0.205},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-4.7-Flash":{"id":"zai-org/GLM-4.7-Flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0.06,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"cost":{"input":0.4,"output":1.75,"cache_read":0.08},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768},"cost":{"input":0.75,"output":2.4,"cache_read":0.14,"cache_write":0.9375},"canonical_model_id":"zhipuai/glm-5.2"},"ByteDance/Seed-2.0-pro":{"id":"ByteDance/Seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.5,"output":3,"cache_read":0.1,"tiers":[{"input":1,"output":6,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"ByteDance/Seed-2.0-code":{"id":"ByteDance/Seed-2.0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1,"tiers":[{"input":1,"output":6,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"ByteDance/Seed-2.0-mini":{"id":"ByteDance/Seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02,"tiers":[{"input":0.2,"output":0.8,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"}}},"standardcompute":{"id":"standardcompute","env":["STANDARDCOMPUTE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.stdcmpt.com/v1","name":"Standard Compute","doc":"https://docs.standardcompute.com/models","models":{"standardcompute":{"id":"standardcompute","name":"Standard Compute","description":"Flat-rate smart-routing gateway: one model id, each request routed across the customer's enabled models from a curated ~1M-context catalog (Claude, GPT, Gemini, DeepSeek, GLM, Qwen, Kimi, MiniMax, Grok, MiMo) or pinned to a user-selected model","family":"auto","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-03-01","last_updated":"2026-10-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"}},"mistralai/mistral-large-4-0":{"id":"mistralai/mistral-large-4-0","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.5625,"output":1.8,"cache_read":0.105},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.08,"output":0.5,"cache_read":0.012},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":0.375,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.083333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.14,"output":0.42,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.044,"output":2.088,"cache_read":0.2088},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.122,"output":3.366,"cache_read":0.1122},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.0826,"output":0.1652,"cache_read":0.01652},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.32,"output":1.28,"cache_read":0.064,"cache_write":0.4,"tiers":[{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.96,"output":3.84,"cache_read":0.192,"cache_write":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.15,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.425,"output":2.55,"cache_read":0.085,"cache_write":0.53125},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":50000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"cache_write":0.25,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":850000,"output":30000},"provider":{"npm":"@openrouter/ai-sdk-provider"},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"}}},"routing-run":{"id":"routing-run","env":["ROUTING_RUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.routing.run/v1","name":"routing.run","doc":"https://docs.routing.run/api-reference/models","models":{"nemotron-3-ultra":{"id":"nemotron-3-ultra","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32000},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32000},"cost":{"input":0.16,"output":0.48},"canonical_model_id":"alibaba/qwen3.5-9b"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.8,"output":2.4},"canonical_model_id":"zhipuai/glm-5.2"},"glm-5.2-nitro":{"id":"glm-5.2-nitro","name":"GLM 5.2 Nitro","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.8,"output":2.4},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.348,"output":0.696},"canonical_model_id":"deepseek/deepseek-v4-pro"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"kimi-k2.7-code-nitro":{"id":"kimi-k2.7-code-nitro","name":"Kimi K2.7 Code Nitro","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.6"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":1.5,"output":9},"canonical_model_id":"openai/gpt-5.6-terra"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":0.7,"output":4.2},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":64000},"cost":{"input":0.112,"output":0.224},"canonical_model_id":"deepseek/deepseek-v4-flash"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15},"canonical_model_id":"openai/gpt-5.6-sol"},"kimi-k2.6-nitro":{"id":"kimi-k2.6-nitro","name":"Kimi K2.6 Nitro","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32000},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"moonshotai/kimi-k2.6"}}},"meganova":{"id":"meganova","env":["MEGANOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.meganova.ai/v1","name":"Meganova","doc":"https://docs.meganova.ai","models":{"mistralai/Mistral-Small-3.2-24B-Instruct-2506":{"id":"mistralai/Mistral-Small-3.2-24B-Instruct-2506","name":"Mistral Small 3.2 24B Instruct","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"mistral/mistral-small-2506"},"mistralai/Mistral-Nemo-Instruct-2407":{"id":"mistralai/Mistral-Nemo-Instruct-2407","name":"Mistral Nemo Instruct 2407","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.02,"output":0.04},"canonical_model_id":"mistral/mistral-nemo"},"MiniMaxAI/MiniMax-M2.1":{"id":"MiniMaxAI/MiniMax-M2.1","name":"MiniMax M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"output":131072},"cost":{"input":0.28,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.1"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.15},"canonical_model_id":"minimax/MiniMax-M2.5"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"moonshotai/Kimi-K2-Thinking":{"id":"moonshotai/Kimi-K2-Thinking","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":2.6},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2026-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.09,"output":0.1},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen2.5-VL-32B-Instruct":{"id":"Qwen/Qwen2.5-VL-32B-Instruct","name":"Qwen2.5 VL 32B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.2,"output":0.6}},"Qwen/Qwen3.5-Plus":{"id":"Qwen/Qwen3.5-Plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":2.4,"reasoning":2.4},"canonical_model_id":"alibaba/qwen3.5-plus"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.19,"output":0.79},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3-0324":{"id":"deepseek-ai/DeepSeek-V3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.275,"output":1.1},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-V3.2-Exp":{"id":"deepseek-ai/DeepSeek-V3.2-Exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-10","last_updated":"2025-10-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.26,"output":0.38,"cache_read":0.13}},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.26,"output":0.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":64000},"cost":{"input":0.5,"output":2.15,"cache_read":0.4},"canonical_model_id":"deepseek/deepseek-r1-0528"},"XiaomiMiMo/MiMo-V2-Flash":{"id":"XiaomiMiMo/MiMo-V2-Flash","name":"MiMo V2 Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32000},"status":"deprecated","cost":{"input":0.1,"output":0.3},"canonical_model_id":"xiaomi/mimo-v2-flash"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.45,"output":1.9},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.8,"output":2.56,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.4,"output":1.75,"cache_read":0.08},"canonical_model_id":"zhipuai/glm-4.7"}}},"chutes":{"id":"chutes","env":["CHUTES_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://llm.chutes.ai/v1","name":"Chutes","doc":"https://llm.chutes.ai/v1/models","models":{"Nemotron-3-Nano-Omni-30B-TEE":{"id":"Nemotron-3-Nano-Omni-30B-TEE","name":"Nemotron 3 Nano Omni 30B TEE","description":"Omni-modal model for text, vision, audio, and multimodal agent tasks","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":0},"cost":{"input":0.0245,"output":0.0978,"cache_read":0.0024499999999999995}},"google/gemma-4-31B-turbo-TEE":{"id":"google/gemma-4-31B-turbo-TEE","name":"gemma 4 31B turbo TEE","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":0.12,"output":0.37,"cache_read":0.011999999999999997},"canonical_model_id":"google/gemma-4-31b-it"},"moonshotai/Kimi-K3-TEE":{"id":"moonshotai/Kimi-K3-TEE","name":"Kimi K3 TEE","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65535},"cost":{"input":1.85,"output":9.5,"cache_read":0.18499999999999997},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.6-TEE":{"id":"moonshotai/Kimi-K2.6-TEE","name":"Kimi K2.6 TEE","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65535},"cost":{"input":0.45,"output":1.95,"cache_read":0.04499999999999999},"canonical_model_id":"moonshotai/kimi-k2.6"},"Qwen/Qwen3-32B-TEE":{"id":"Qwen/Qwen3-32B-TEE","name":"Qwen3 32B TEE","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":40960},"cost":{"input":0.104,"output":0.416,"cache_read":0.010399999999999998},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B-TEE":{"id":"Qwen/Qwen3.6-27B-TEE","name":"Qwen3.6 27B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.024999999999999994},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3.8-27B-TEE":{"id":"Qwen/Qwen3.8-27B-TEE","name":"Qwen3.8 27B TEE","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.039,"output":1.25,"cache_read":0.003899999999999999},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3-235B-A22B-Thinking-2507-TEE":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507-TEE","name":"Qwen3 235B A22B Thinking 2507 TEE","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2989,"output":1.1957,"cache_read":0.029889999999999993},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen3.5-397B-A17B-TEE":{"id":"Qwen/Qwen3.5-397B-A17B-TEE","name":"Qwen3.5 397B A17B TEE","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.45,"output":3,"cache_read":0.04499999999999999},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"unsloth/Mistral-Nemo-Instruct-2407-TEE":{"id":"unsloth/Mistral-Nemo-Instruct-2407-TEE","name":"Mistral Nemo Instruct 2407 TEE","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0245,"output":0.0978,"cache_read":0.0024499999999999995},"canonical_model_id":"mistral/mistral-nemo"},"deepseek-ai/DeepSeek-V4-Flash-0731-TEE":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731-TEE","name":"DeepSeek V4 Flash 0731 TEE","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.44,"output":1.32,"cache_read":0.04399999999999999},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V3.2-TEE":{"id":"deepseek-ai/DeepSeek-V3.2-TEE","name":"DeepSeek V3.2 TEE","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":65536},"cost":{"input":1,"output":1,"cache_read":0.09999999999999998},"canonical_model_id":"deepseek/deepseek-v3.2"},"zai-org/GLM-5.1-TEE":{"id":"zai-org/GLM-5.1-TEE","name":"GLM 5.1 TEE","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":65535},"cost":{"input":0.98,"output":3.08,"cache_read":0.09799999999999998},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5.2-TEE":{"id":"zai-org/GLM-5.2-TEE","name":"GLM 5.2 TEE","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65535},"cost":{"input":1.25,"output":3.95,"cache_read":0.12499999999999997},"canonical_model_id":"zhipuai/glm-5.2"}}},"drun":{"id":"drun","env":["DRUN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://chat.d.run/v1","name":"D.Run (China)","doc":"https://www.d.run","models":{"public/minimax-m25":{"id":"public/minimax-m25","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.29,"output":1.16},"canonical_model_id":"minimax/MiniMax-M2.5"},"public/deepseek-v3":{"id":"public/deepseek-v3","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.28,"output":1.1},"canonical_model_id":"deepseek/deepseek-v3"},"public/deepseek-r1":{"id":"public/deepseek-r1","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32000},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"deepseek/deepseek-r1"}}},"io-net":{"id":"io-net","env":["IOINTELLIGENCE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.intelligence.io.solutions/api/v1","name":"IO.NET","doc":"https://io.net/docs/guides/intelligence/io-intelligence","models":{"mistralai/Magistral-Small-2506":{"id":"mistralai/Magistral-Small-2506","name":"Magistral Small 2506","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.5,"output":1.5,"cache_read":0.25,"cache_write":1},"canonical_model_id":"mistral/magistral-small-2506"},"mistralai/Mistral-Nemo-Instruct-2407":{"id":"mistralai/Mistral-Nemo-Instruct-2407","name":"Mistral Nemo Instruct 2407","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-05","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.0514,"output":0.082,"cache_read":0.0257},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/Devstral-Small-2505":{"id":"mistralai/Devstral-Small-2505","name":"Devstral Small 2505","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-05-01","last_updated":"2025-05-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.05,"output":0.22,"cache_read":0.025,"cache_write":0.1}},"mistralai/Mistral-Large-Instruct-2411":{"id":"mistralai/Mistral-Large-Instruct-2411","name":"Mistral Large Instruct 2411","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-11-18","last_updated":"2024-11-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":2,"output":6,"cache_read":1,"cache_write":4},"canonical_model_id":"mistral/mistral-large-2411"},"meta-llama/Llama-3.2-90B-Vision-Instruct":{"id":"meta-llama/Llama-3.2-90B-Vision-Instruct","name":"Llama 3.2 90B Vision Instruct","description":"Open Llama multimodal model for image understanding and text reasoning","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4096},"status":"deprecated","cost":{"input":0.35,"output":0.4,"cache_read":0.175,"cache_write":0.7}},"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"id":"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","name":"Llama 4 Maverick 17B 128E Instruct","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":430000,"output":4096},"cost":{"input":0.26875,"output":0.82625,"cache_read":0.134375},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.5466,"output":1.0226,"cache_read":0.2733},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"moonshotai/Kimi-K2-Instruct-0905":{"id":"moonshotai/Kimi-K2-Instruct-0905","name":"Kimi K2 Instruct","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":0.57,"output":2.3,"cache_read":0.285},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/Kimi-K2-Thinking":{"id":"moonshotai/Kimi-K2-Thinking","name":"Kimi K2 Thinking","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096},"cost":{"input":0.6,"output":2.5,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"Qwen/Qwen3-235B-A22B-Thinking-2507":{"id":"Qwen/Qwen3-235B-A22B-Thinking-2507","name":"Qwen 3 235B Thinking","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"status":"deprecated","cost":{"input":0.11,"output":0.6,"cache_read":0.055,"cache_write":0.22},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"Qwen/Qwen2.5-VL-32B-Instruct":{"id":"Qwen/Qwen2.5-VL-32B-Instruct","name":"Qwen 2.5 VL 32B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-11-01","last_updated":"2024-11-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"status":"deprecated","cost":{"input":0.05,"output":0.22,"cache_read":0.025,"cache_write":0.1}},"Qwen/Qwen3-Next-80B-A3B-Instruct":{"id":"Qwen/Qwen3-Next-80B-A3B-Instruct","name":"Qwen 3 Next 80B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"cost":{"input":0.11875,"output":1.2625,"cache_read":0.059375},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.523667,"output":2.205333,"cache_read":0.261834},"canonical_model_id":"deepseek/deepseek-r1-0528"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT-OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":0.188,"output":0.7,"cache_read":0.094},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT-OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":4096},"cost":{"input":0.055,"output":0.194,"cache_read":0.0275},"canonical_model_id":"openai/gpt-oss-20b"},"Intel/Qwen3-Coder-480B-A35B-Instruct-int4-mixed-ar":{"id":"Intel/Qwen3-Coder-480B-A35B-Instruct-int4-mixed-ar","name":"Qwen 3 Coder 480B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":106000,"output":4096},"cost":{"input":0.46125,"output":2.29375,"cache_read":0.230625},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM 4.6","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":4096},"cost":{"input":0.52,"output":2.0375,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-4.6"}}},"alibaba-coding-plan":{"id":"alibaba-coding-plan","env":["ALIBABA_CODING_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://coding-intl.dashscope.aliyuncs.com/v1","name":"Alibaba Coding Plan","doc":"https://www.alibabacloud.com/help/en/model-studio/coding-plan","models":{"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"alibaba/qwen3.7-plus"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":16384},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"glm-5":{"id":"glm-5","name":"GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-11","last_updated":"2026-02-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202752,"output":16384},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-coder-plus":{"id":"qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"qwen3-max-2026-01-23":{"id":"qwen3-max-2026-01-23","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}},"MiniMax-M2.5":{"id":"MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196608,"input":196601,"output":24576},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0}}}},"requesty":{"id":"requesty","env":["REQUESTY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://router.requesty.ai/v1","name":"Requesty","doc":"https://requesty.ai/solution/llm-routing/models","models":{"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.43,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"claude-sonnet-5@eu":{"id":"claude-sonnet-5@eu","name":"Claude Sonnet 5 (EU)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5"},"muse-wonder-1.0":{"id":"muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"meta/muse-wonder-1.0"},"nemotron-3-nano-omni":{"id":"nemotron-3-nano-omni","name":"nemotron-3-nano-omni","description":"The most open, efficient, and accurate omni modal reasoning model for agentic AI.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-20","last_updated":"2026-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":300000},"cost":{"input":0.06,"output":0.24,"cache_read":0.06}},"deepseek-v4.1-flash-1017@eu":{"id":"deepseek-v4.1-flash-1017@eu","name":"DeepSeek V4.1 Flash 1017 (EU)","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":2},"canonical_model_id":"google/gemini-3.1-flash-image"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"claude-haiku-4-5@eu":{"id":"claude-haiku-4-5@eu","name":"Claude Haiku 4.5 (latest) (EU)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"gemini-3.5-flash-lite@eu":{"id":"gemini-3.5-flash-lite@eu","name":"Gemini 3.5 Flash Lite (EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.33,"output":2.75,"cache_read":0.033},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"devstral-latest":{"id":"devstral-latest","name":"devstral-latest","description":"An enterprise grade text model, that excels at using tools to explore codebases, editing multiple files and power software engineering agents.","family":"devstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.44,"output":2.2,"cache_read":0.44}},"gpt-5-nano@eu":{"id":"gpt-5-nano@eu","name":"GPT-5 Nano (EU)","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":0.055,"output":0.44,"cache_read":0.0055},"canonical_model_id":"openai/gpt-5-nano"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.8,"output":2.55,"cache_read":0.16},"canonical_model_id":"zhipuai/glm-5.2"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.13,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.63,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"minimax-m2.7-highspeed":{"id":"minimax-m2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.6,"output":2.4,"cache_read":0.06,"cache_write":1.2},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"gpt-4.1-nano@eu":{"id":"gpt-4.1-nano@eu","name":"GPT-4.1 nano (EU)","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.11,"output":0.44,"cache_read":0.0275},"canonical_model_id":"openai/gpt-4.1-nano"},"glm-5.3-flash@eu":{"id":"glm-5.3-flash@eu","name":"GLM-5.3-Flash (EU)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0.2,"output":0.6,"cache_read":0.07},"canonical_model_id":"zhipuai/glm-5.3-flash"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.32,"output":1.28,"cache_read":0.032,"cache_write":0.4},"canonical_model_id":"alibaba/qwen3.7-plus"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"kimi-k3.1-pro":{"id":"kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"qwen3.8-2.4T-A95B":{"id":"qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":1.583},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.75,"output":16.5,"cache_read":0.275},"canonical_model_id":"openai/gpt-5.4"},"qwen3.8-flash-next@eu":{"id":"qwen3.8-flash-next@eu","name":"Qwen3.8 Flash Next (EU)","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"claude-sonnet-4@eu":{"id":"claude-sonnet-4@eu","name":"Claude Sonnet 4 (latest) (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"claude-opus-4-6@eu":{"id":"claude-opus-4-6@eu","name":"Claude Opus 4.6 (EU)","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-6"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"gpt-6.1-sol@eu":{"id":"gpt-6.1-sol@eu","name":"GPT-6.1 Sol (EU)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":12,"cache_read":0.12,"cache_write":3,"tiers":[{"input":4.8,"output":18,"cache_read":0.24,"cache_write":6,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.8,"output":18,"cache_read":0.24,"cache_write":6}},"canonical_model_id":"openai/gpt-6.1-sol"},"laguna-xs.2":{"id":"laguna-xs.2","name":"Laguna XS.2","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-xs.2"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"seed-2.0-pro":{"id":"seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"o4-mini@eu":{"id":"o4-mini@eu","name":"o4-mini (EU)","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.21,"output":4.84,"cache_read":0.3025},"canonical_model_id":"openai/o4-mini"},"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"gpt-5@eu":{"id":"gpt-5@eu","name":"GPT-5 (EU)","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":11,"cache_read":0.1375},"canonical_model_id":"openai/gpt-5"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":4.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":9,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":9}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":20480},"cost":{"input":0,"output":0},"canonical_model_id":"meta/muse-glimmer-30b"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro"},"mistral-large-4@eu":{"id":"mistral-large-4@eu","name":"Mistral Large 4 (EU)","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"laguna-m.1":{"id":"laguna-m.1","name":"Laguna M.1","description":"Poolside's open-weight model for agentic coding and long-horizon work","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"poolside/laguna-m.1"},"nemotron-3.5-lightning-30b-a3b":{"id":"nemotron-3.5-lightning-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"gpt-5.4@eu":{"id":"gpt-5.4@eu","name":"GPT-5.4 (EU)","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.75,"output":16.5,"cache_read":0.275,"tiers":[{"input":5.5,"output":24.75,"cache_read":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5.5,"output":24.75,"cache_read":0.55}},"canonical_model_id":"openai/gpt-5.4"},"claude-fable-5.1@eu":{"id":"claude-fable-5.1@eu","name":"Claude Fable 5.1 (EU)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":0.275,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-6-sol@eu":{"id":"gpt-6-sol@eu","name":"GPT-6 Sol (EU)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":12,"cache_read":0.24,"cache_write":3,"tiers":[{"input":4.8,"output":18,"cache_read":0.48,"cache_write":6,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.8,"output":18,"cache_read":0.48,"cache_write":6}},"canonical_model_id":"openai/gpt-6-sol"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"deepseek-v4-flash-0731@eu":{"id":"deepseek-v4-flash-0731@eu","name":"DeepSeek V4 Flash 0731 (EU)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gemini-2.5-flash@eu":{"id":"gemini-2.5-flash@eu","name":"Gemini 2.5 Flash (EU)","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.075,"cache_write":0.55},"canonical_model_id":"google/gemini-2.5-flash"},"mistral-small-2603@eu":{"id":"mistral-small-2603@eu","name":"Mistral Small 4 (EU)","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.165,"output":0.66,"cache_read":0.165},"canonical_model_id":"mistral/mistral-small-2603"},"inkling-256k":{"id":"inkling-256k","name":"Inkling 256K","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"ring-2.6-1t":{"id":"ring-2.6-1t","name":"ring-2.6-1t","description":"Inclusion AI ring-2.6-1t","family":"ring","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-08","last_updated":"2026-05-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.5}},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2},"canonical_model_id":"xai/grok-4.7"},"nemotron-3-ultra-nvfp4":{"id":"nemotron-3-ultra-nvfp4","name":"nemotron-3-ultra-nvfp4","description":"Nemotron-3-Ultra-550B-A55B-NVFP4 is a frontier-scale large language model (LLM) trained by NVIDIA, designed to deliver strong agentic, reasoning, and conversational capabilities. It is optimized for the most demanding workloads, including complex multi-step agents, long-context analysis, and high-accuracy reasoning over code, math, and science. The model employs a hybrid Latent Mixture-of-Experts (LatentMoE) architecture, utilizing interleaved Mamba-2 and MoE layers, along with select Attention layers. Like the Super model, the Ultra model incorporates Multi-Token Prediction (MTP) layers for faster text generation and improved quality, and it is trained using an NVFP4 pre-training recipe to maximize compute efficiency. The model has 55B active parameters and 550B parameters in total.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":2.4,"cache_read":0.12}},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-flash-next":{"id":"qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.6"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5.5,"output":33,"cache_read":0.55},"canonical_model_id":"openai/gpt-5.5"},"qwen3.8-2.4T-A95B@eu":{"id":"qwen3.8-2.4T-A95B@eu","name":"Qwen3.8 2.4T A95B (EU)","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":2.5,"output":6,"cache_read":0.63},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"seed-1.8":{"id":"seed-1.8","name":"Seed 1.8","description":"ByteDance Seed model for multimodal reasoning, long-context analysis, and agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":0.25,"output":2,"cache_read":0.05},"canonical_model_id":"bytedance-seed/seed-1-8"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-sonnet-4-5@eu":{"id":"claude-sonnet-4-5@eu","name":"Claude Sonnet 4.5 (latest) (EU)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.3,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.6,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.6,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"claude-sonnet-5-5@eu":{"id":"claude-sonnet-5-5@eu","name":"Claude Sonnet 5.5 (EU)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.22,"cache_write":2.75},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"kimi-k3@eu":{"id":"kimi-k3@eu","name":"Kimi K3 (EU)","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":3,"output":15,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-v4-pro-0813@eu":{"id":"deepseek-v4-pro-0813@eu","name":"DeepSeek V4 Pro 0813 (EU)","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":1.75,"output":3.5,"cache_read":0.44},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"xiaomi/mimo-v2.5"},"glm-5.1@eu":{"id":"glm-5.1@eu","name":"GLM-5.1 (EU)","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.1"},"gemini-3.5-flash@eu":{"id":"gemini-3.5-flash@eu","name":"Gemini 3.5 Flash (EU)","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.65,"output":9.9,"cache_read":0.165,"cache_write":1.7413},"canonical_model_id":"google/gemini-3.5-flash"},"ling-3.0-tiny":{"id":"ling-3.0-tiny","name":"ling-3.0-tiny","description":"Ling-3.0-tiny is an efficient 7.9B parameter MoE model from inclusionAI with only 1.3B active parameters per token. Built for responsive agents, reliable instruction following and multi turn conversation, with a 256K context window, native function calling, prompt caching and switchable Thinking and Instant modes.","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"gemini-2.5-pro@eu":{"id":"gemini-2.5-pro@eu","name":"Gemini 2.5 Pro (EU)","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.25,"output":10,"cache_read":0.31,"cache_write":2.375,"tiers":[{"input":2.5,"output":15,"cache_read":0.62,"cache_write":4.75,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.62,"cache_write":4.75}},"canonical_model_id":"google/gemini-2.5-pro"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"cache_write":1.25,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4,"cache_write":2.5}},"canonical_model_id":"xai/grok-4.3"},"ling-2.6-flash":{"id":"ling-2.6-flash","name":"ling-2.6-flash","description":"Inclusion AI ling-2.6-flash","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3}},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":4.5},"canonical_model_id":"google/gemini-3-pro-image"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"gpt-5.6-sol@eu":{"id":"gpt-5.6-sol@eu","name":"GPT-5.6 Sol (EU)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.44,"cache_write":5.5,"tiers":[{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8.8,"output":33,"cache_read":0.88,"cache_write":11}},"canonical_model_id":"openai/gpt-5.6-sol"},"gpt-5.5@eu":{"id":"gpt-5.5@eu","name":"GPT-5.5 (EU)","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5.5,"output":33,"cache_read":0.55,"tiers":[{"input":11,"output":49.5,"cache_read":1.1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":11,"output":49.5,"cache_read":1.1}},"canonical_model_id":"openai/gpt-5.5"},"minimax-m3@eu":{"id":"minimax-m3@eu","name":"MiniMax-M3 (EU)","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.4,"output":2,"cache_read":0.1},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":0.625},"canonical_model_id":"alibaba/qwen3.6-plus"},"leanstral-1-5":{"id":"leanstral-1-5","name":"leanstral-1-5","description":"Leanstral 1.5 is an updated Lean 4 formal proof engineering model from Mistral AI, optimized for automated theorem proving and autoformalization. It has 119B total parameters with 6.5B active and supports a 256K token context window. It supports native function calling and structured output.","family":"leanstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180},"canonical_model_id":"openai/gpt-5.5-pro"},"claude-fable-5.1":{"id":"claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262144},"cost":{"input":3,"output":15,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"nemotron-3.5-content-safety":{"id":"nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"kimi-k2.6@eu":{"id":"kimi-k2.6@eu","name":"Kimi K2.6 (EU)","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"cost":{"input":0.95,"output":4,"cache_read":0.95},"canonical_model_id":"moonshotai/kimi-k2.6"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"gpt-4o-mini@eu":{"id":"gpt-4o-mini@eu","name":"GPT-4o mini (EU)","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.165,"output":0.66,"cache_read":0.0825},"canonical_model_id":"openai/gpt-4o-mini"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"glm-5.2@eu":{"id":"glm-5.2@eu","name":"GLM-5.2 (EU)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.2,"output":4.2,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-4.1-mini@eu":{"id":"gpt-4.1-mini@eu","name":"GPT-4.1 mini (EU)","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.44,"output":1.76,"cache_read":0.11},"canonical_model_id":"openai/gpt-4.1-mini"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0027},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"claude-opus-5@eu":{"id":"claude-opus-5@eu","name":"Claude Opus 5 (EU)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-5"},"claude-haiku-5-5@eu":{"id":"claude-haiku-5-5@eu","name":"Claude Haiku 5.5 (EU)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.11,"output":0.55,"cache_read":0.02,"cache_write":0.14,"tiers":[{"input":0.55,"output":2.75,"cache_read":0.06,"cache_write":0.69,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"seed-2.0-mini":{"id":"seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.4-pro":{"id":"gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"cache_read":30},"canonical_model_id":"openai/gpt-5.4-pro"},"glm-5.2-fast":{"id":"glm-5.2-fast","name":"glm-5.2-fast","description":"GLM-5.2 introduces a robust 1M-token context and advanced, multi-effort coding capabilities to significantly enhance performance on long-horizon tasks. Its new IndexShare architecture and improved MTP layer simultaneously boost efficiency by reducing per-token FLOPs and increasing speculative decoding lengths. A 743B-parameter model in Zhipu AI's GLM series, designed to plan, execute, and iterate autonomously on extended, engineering-grade tasks.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-07-13","last_updated":"2026-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.1,"output":6.6,"cache_read":0.21}},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"nvidia-nemotron-3-ultra":{"id":"nvidia-nemotron-3-ultra","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.5,"output":2.5},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"gpt-5.6-terra@eu":{"id":"gpt-5.6-terra@eu","name":"GPT-5.6 Terra (EU)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.2,"cache_read":0.22,"cache_write":2.75,"tiers":[{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4.4,"output":19.8,"cache_read":0.44,"cache_write":5.5}},"canonical_model_id":"openai/gpt-5.6-terra"},"mistral-large-4":{"id":"mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"gpt-6-luna@eu":{"id":"gpt-6-luna@eu","name":"GPT-6 Luna (EU)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.12,"output":0.6,"cache_read":0.012,"cache_write":0.15,"tiers":[{"input":0.24,"output":0.9,"cache_read":0.024,"cache_write":0.3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.24,"output":0.9,"cache_read":0.024,"cache_write":0.3}},"canonical_model_id":"openai/gpt-6-luna"},"gemini-3.7-flash@eu":{"id":"gemini-3.7-flash@eu","name":"Gemini 3.7 Flash (EU)","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.825,"output":4.125,"cache_read":0.0825},"canonical_model_id":"google/gemini-3.7-flash"},"claude-sonnet-4-6@eu":{"id":"claude-sonnet-4-6@eu","name":"Claude Sonnet 4.6 (EU)","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.3,"cache_write":4.125},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"ling-2.6-1t":{"id":"ling-2.6-1t","name":"ling-2.6-1t","description":"Inclusion AI ling-2.6-1t","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.5}},"claude-opus-4-7@eu":{"id":"claude-opus-4-7@eu","name":"Claude Opus 4.7 (EU)","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-5-mini@eu":{"id":"gpt-5-mini@eu","name":"GPT-5 Mini (EU)","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":0.275,"output":2.2,"cache_read":0.0275},"canonical_model_id":"openai/gpt-5-mini"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.2,"output":4.2,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"claude-opus-5-5@eu":{"id":"claude-opus-5-5@eu","name":"Claude Opus 5.5 (EU)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.4,"output":22,"cache_read":0.22,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"deepseek-v4-pro@eu":{"id":"deepseek-v4-pro@eu","name":"DeepSeek V4 Pro (EU)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.75,"output":3.5,"cache_read":0.44},"canonical_model_id":"deepseek/deepseek-v4-pro"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"alibaba/qwen3.7-max"},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.1},"canonical_model_id":"xai/grok-build-0.1"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.58,"cache_read":0.035},"canonical_model_id":"tencent/hy3"},"nemotron-lightning-3.5-30b-a3b":{"id":"nemotron-lightning-3.5-30b-a3b","name":"nemotron-lightning-3.5-30b-a3b","description":"Nemotron-Lightning-3.5-30B-A3B is a 30B-parameter Mixture-of-Experts language model (3B active) from NVIDIA's Nemotron-H family, built on a hybrid Mamba-Transformer architecture for efficient long-context inference. Like other models in the family, it responds to queries by first generating a reasoning trace and then concluding with a final response, with reasoning behavior configurable through a flag in the chat template. It includes a multi-token prediction (MTP) speculative decoding head for low-latency serving.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-15","last_updated":"2026-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":262144},"cost":{"input":0.05,"output":0.2,"cache_read":0.01}},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.5,"output":7,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"nemotron-3-super-120b-a12b":{"id":"nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"grok-4.2-beta":{"id":"grok-4.2-beta","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":0.4,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.4,"cache_write":4}},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"leanstral-1-5@eu":{"id":"leanstral-1-5@eu","name":"leanstral-1-5@eu","description":"Leanstral 1.5 is an updated Lean 4 formal proof engineering model from Mistral AI, optimized for automated theorem proving and autoformalization. It has 119B total parameters with 6.5B active and supports a 256K token context window. It supports native function calling and structured output.","family":"leanstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0}},"kat-coder-pro":{"id":"kat-coder-pro","name":"kat-coder-pro","description":"KAT-Coder-Pro V2 by KwaiKAT is a non-reasoning model optimized for agentic coding. It delivers strong performance on reasoning-style tasks while requiring significantly fewer output tokens than peer models. With the 1210 release, it achieved a score of 64 on the Artificial Analysis Intelligence Index, placing it in the global Top 10 and ranking first among all non-reasoning models.","family":"kat-coder","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-03-27","last_updated":"2026-03-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.3,"output":1.2}},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"gemini-2.5-flash-lite@eu":{"id":"gemini-2.5-flash-lite@eu","name":"Gemini 2.5 Flash-Lite (EU)","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"cache_write":0.18333},"canonical_model_id":"google/gemini-2.5-flash-lite"},"nvidia-nemotron-3-super-120b-a12b":{"id":"nvidia-nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.5},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"seed-2.0-code":{"id":"seed-2.0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.1},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"gpt-4.1@eu":{"id":"gpt-4.1@eu","name":"GPT-4.1 (EU)","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2.2,"output":8.8,"cache_read":0.55},"canonical_model_id":"openai/gpt-4.1"},"mistral-medium-latest":{"id":"mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.44,"output":2.2,"cache_read":0.44},"canonical_model_id":"mistral/mistral-medium-latest"},"fugu-ultra":{"id":"fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.165,"output":0.66,"cache_read":0.165},"canonical_model_id":"mistral/mistral-small-2603"},"gemini-3.8-flash@eu":{"id":"gemini-3.8-flash@eu","name":"Gemini 3.8 Flash (EU)","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.825,"output":4.125,"cache_read":0.0825},"canonical_model_id":"google/gemini-3.8-flash"},"qwen3.5-2b":{"id":"qwen3.5-2b","name":"Qwen3.5 2B","description":"Compact Qwen3.5 multimodal model for fast chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.02,"output":0.1},"canonical_model_id":"alibaba/qwen3.5-2b"},"nemotron-3-ultra-550b-a55b":{"id":"nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"gemini-4-argon@eu":{"id":"gemini-4-argon@eu","name":"Gemini 4 Argon (EU)","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2.2,"output":11,"cache_read":0.11},"canonical_model_id":"google/gemini-4-argon"},"gpt-5.1@eu":{"id":"gpt-5.1@eu","name":"GPT-5.1 (EU)","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":11,"cache_read":0.1375},"canonical_model_id":"openai/gpt-5.1"},"kimi-k3.1-pro-highspeed":{"id":"kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.07,"output":0.34,"cache_read":0.07},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"mistral-medium-3-5@eu":{"id":"mistral-medium-3-5@eu","name":"Mistral Medium 3.5 (EU)","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.65,"output":8.25,"cache_read":1.65},"canonical_model_id":"mistral/mistral-medium-3-5"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.28,"output":0.56,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":262144},"cost":{"input":0.2,"output":0.6,"cache_read":0.07},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gemini-3.1-flash-lite@eu":{"id":"gemini-3.1-flash-lite@eu","name":"Gemini 3.1 Flash Lite (EU)","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.275,"output":1.65,"cache_read":0.0275,"cache_write":0.091663},"canonical_model_id":"google/gemini-3.1-flash-lite"},"kimi-k2.7-code@eu":{"id":"kimi-k2.7-code@eu","name":"Kimi K2.7 Code (EU)","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.25,"output":4.5,"cache_read":0.31},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"nemotron-3-nano-omni@eu":{"id":"nemotron-3-nano-omni@eu","name":"nemotron-3-nano-omni@eu","description":"The most open, efficient, and accurate omni modal reasoning model for agentic AI.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-20","last_updated":"2026-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":300000},"cost":{"input":0.06,"output":0.24,"cache_read":0.06}},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"mistral-medium-latest@eu":{"id":"mistral-medium-latest@eu","name":"Mistral Medium (latest) (EU)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.44,"output":2.2,"cache_read":0.44},"canonical_model_id":"mistral/mistral-medium-latest"},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.26,"output":2.6},"canonical_model_id":"alibaba/qwen3.5-27b"},"nemotron-3-nano-omni-30b-a3b-reasoning":{"id":"nemotron-3-nano-omni-30b-a3b-reasoning","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":20480},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"deepseek-v4.1-flash@eu":{"id":"deepseek-v4.1-flash@eu","name":"DeepSeek V4.1 Flash (EU)","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"devstral-latest@eu":{"id":"devstral-latest@eu","name":"devstral-latest@eu","description":"An enterprise grade text model, that excels at using tools to explore codebases, editing multiple files and power software engineering agents.","family":"devstral","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-27","last_updated":"2026-05-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.44,"output":2.2,"cache_read":0.44}},"inkling":{"id":"inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":32768},"cost":{"input":1.87,"output":4.68,"cache_read":0.374},"canonical_model_id":"thinkingmachines/inkling"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"cache_write":4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1,"cache_write":4}},"canonical_model_id":"xai/grok-4.5"},"claude-opus-4-5@eu":{"id":"claude-opus-4-5@eu","name":"Claude Opus 4.5 (latest) (EU)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-5"},"glm-5.3@eu":{"id":"glm-5.3@eu","name":"GLM-5.3 (EU)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.2,"output":4.2,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"mistral-medium-3-5":{"id":"mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715},"cost":{"input":1.65,"output":8.25,"cache_read":1.65},"canonical_model_id":"mistral/mistral-medium-3-5"},"gpt-5.6-luna@eu":{"id":"gpt-5.6-luna@eu","name":"GPT-5.6 Luna (EU)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.22,"output":1.32,"cache_read":0.022,"cache_write":0.275,"tiers":[{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.44,"output":1.98,"cache_read":0.044,"cache_write":0.55}},"canonical_model_id":"openai/gpt-5.6-luna"},"kimi-k2.5@eu":{"id":"kimi-k2.5@eu","name":"Kimi K2.5 (EU)","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16000},"cost":{"input":0.72,"output":3.6,"cache_read":0.72,"cache_write":0.72},"canonical_model_id":"moonshotai/kimi-k2.5"},"claude-opus-4-8@eu":{"id":"claude-opus-4-8@eu","name":"Claude Opus 4.8 (EU)","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.5,"cache_read":0.55,"cache_write":6.875},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-fable-5@eu":{"id":"claude-fable-5@eu","name":"Claude Fable 5 (EU)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"muse-spark-open":{"id":"muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"}}},"neosmith":{"id":"neosmith","env":["NEOSMITH_API_KEY"],"npm":"@ai-sdk/openai","api":"https://router.neosmith.ai/v1","name":"NeoSmith","doc":"https://neosmith.ai/docs","models":{"neosmith.intelligent-basic":{"id":"neosmith.intelligent-basic","name":"NeoSmith Basic","description":"Cost-capped tier. Intelligent routing with a Claude Sonnet ceiling — Opus is never invoked.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-26","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.17,"output":4.37,"cache_read":0.22,"cache_write":0}},"neosmith.intelligent-pro":{"id":"neosmith.intelligent-pro","name":"NeoSmith Pro","description":"Default production tier. Intelligent NeoSmith routing with a Claude Opus ceiling on escalation.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-26","last_updated":"2026-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.81,"output":8.39,"cache_read":0.3,"cache_write":0}},"neosmith.neolite":{"id":"neosmith.neolite","name":"NeoSmith NeoLite","description":"Sealed single-model budget tier. 512K context, text and images, tool use, and no escalation of any kind.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-20","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":512000,"output":64000},"cost":{"input":0.6,"output":2.4,"cache_read":0.08,"cache_write":0}},"neosmith.intelligent-maestro":{"id":"neosmith.intelligent-maestro","name":"NeoSmith Maestro","description":"Highest-accuracy coding tier. Hard, self-contained problems run NeoSmith's premium multi-model solver; everything else gets the strongest intelligence tier.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.4,"output":12,"cache_read":0.35,"cache_write":0}}}},"poe":{"id":"poe","env":["POE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.poe.com/v1","name":"Poe","doc":"https://creator.poe.com/docs/external-applications/openai-compatible-api","models":{"ideogramai/ideogram-v2":{"id":"ideogramai/ideogram-v2","type":"image","name":"Ideogram-v2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-08-21","last_updated":"2024-08-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"ideogramai/ideogram":{"id":"ideogramai/ideogram","type":"image","name":"Ideogram","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-04-03","last_updated":"2024-04-03","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"ideogramai/ideogram-v2a":{"id":"ideogramai/ideogram-v2a","type":"image","name":"Ideogram-v2a","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-02-27","last_updated":"2025-02-27","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"ideogramai/ideogram-v2a-turbo":{"id":"ideogramai/ideogram-v2a-turbo","type":"image","name":"Ideogram-v2a-Turbo","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"ideogram","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-02-27","last_updated":"2025-02-27","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":150,"output":0}},"google/gemma-4-31b":{"id":"google/gemma-4-31b","name":"Gemma-4-31B","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini-3.5-Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5152,"output":9.0909,"cache_read":0.1515},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini-3.8-Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.7576,"output":3.7879,"cache_read":0.0758},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini-3.7-Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.7576,"output":3.7879,"cache_read":0.0758},"canonical_model_id":"google/gemini-3.7-flash"},"google/lyria":{"id":"google/lyria","name":"Lyria","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"lyria","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-06-04","last_updated":"2025-06-04","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0}},"google/imagen-4-ultra":{"id":"google/imagen-4-ultra","type":"image","name":"Imagen-4-Ultra","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-24","last_updated":"2025-05-24","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini-2.5-Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1065535,"output":65535},"status":"deprecated","cost":{"input":0.21,"output":1.8,"cache_read":0.021},"canonical_model_id":"google/gemini-2.5-flash"},"google/imagen-3-fast":{"id":"google/imagen-3-fast","type":"image","name":"Imagen-3-Fast","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-10-17","last_updated":"2024-10-17","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/veo-3.1":{"id":"google/veo-3.1","type":"video","name":"Veo-3.1","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/gemini-deep-research":{"id":"google/gemini-deep-research","name":"gemini-deep-research","description":"Legacy model retained for compatibility with older integrations","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":0},"status":"deprecated","cost":{"input":1.6,"output":9.6}},"google/imagen-4-fast":{"id":"google/imagen-4-fast","type":"image","name":"Imagen-4-Fast","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-06-25","last_updated":"2025-06-25","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini-2.5-Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":32768}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1065535,"output":65535},"status":"deprecated","cost":{"input":0.87,"output":7,"cache_read":0.087},"canonical_model_id":"google/gemini-2.5-pro"},"google/imagen-3":{"id":"google/imagen-3","type":"image","name":"Imagen-3","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-10-15","last_updated":"2024-10-15","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/veo-3-fast":{"id":"google/veo-3-fast","type":"video","name":"Veo-3-Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-13","last_updated":"2025-10-13","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/nano-banana-pro":{"id":"google/nano-banana-pro","type":"image","name":"Nano-Banana-Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":65536,"output":0},"cost":{"input":2.0202,"output":12.1212,"cache_read":0.202},"canonical_model_id":"google/gemini-3-pro-image-preview"},"google/veo-3":{"id":"google/veo-3","type":"video","name":"Veo-3","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-21","last_updated":"2025-05-21","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-omni-1.1-flash":{"id":"google/gemini-omni-1.1-flash","type":"video","name":"Gemini-Omni-1.1-Flash","description":"Video generation and editing model for scene extension, frame interpolation, and reference-guided multimodal control","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":false,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":131072,"output":57920},"cost":{"input":1.5152,"output":9.0909},"canonical_model_id":"google/gemini-omni-1.1-flash"},"google/gemini-2.0-flash":{"id":"google/gemini-2.0-flash","name":"Gemini-2.0-Flash","description":"Earlier Gemini Flash workhorse for responsive multimodal apps and tool use","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":990000,"output":8192},"status":"deprecated","cost":{"input":0.1,"output":0.42},"canonical_model_id":"google/gemini-2.0-flash"},"google/gemini-3.1-pro":{"id":"google/gemini-3.1-pro","name":"Gemini-3.1-Pro","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2.0202,"output":12.1212,"cache_read":0.202},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini-3.6-Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.7576,"output":3.7879,"cache_read":0.0758},"canonical_model_id":"google/gemini-3.6-flash"},"google/imagen-4":{"id":"google/imagen-4","type":"image","name":"Imagen-4","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"imagen","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/gemini-2.0-flash-lite":{"id":"google/gemini-2.0-flash-lite","name":"Gemini-2.0-Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":990000,"output":8192},"status":"deprecated","cost":{"input":0.052,"output":0.21},"canonical_model_id":"google/gemini-2.0-flash-lite"},"google/veo-2":{"id":"google/veo-2","type":"video","name":"Veo-2","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-12-02","last_updated":"2024-12-02","modalities":{"input":["text"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0},"status":"deprecated"},"google/gemini-3-flash":{"id":"google/gemini-3-flash","name":"Gemini-3-Flash","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.404,"output":2.4242,"cache_read":0.0404},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini-3.5-Flash-Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.303,"output":2.5253,"cache_read":0.0303},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini-3.1-Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.2525,"output":1.5152},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini-2.5-Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1024000,"output":64000},"status":"deprecated","cost":{"input":0.07,"output":0.28},"canonical_model_id":"google/gemini-2.5-flash-lite"},"google/veo-3.1-fast":{"id":"google/veo-3.1-fast","type":"video","name":"Veo-3.1-Fast","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":480,"output":0}},"google/nano-banana":{"id":"google/nano-banana","type":"image","name":"Nano-Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":0},"status":"deprecated","cost":{"input":0.21,"output":1.8,"cache_read":0.021},"canonical_model_id":"google/gemini-2.5-flash-image"},"topazlabs-co/topazlabs":{"id":"topazlabs-co/topazlabs","type":"image","name":"TopazLabs","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"topazlabs","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":204,"output":0}},"xai/grok-3-mini":{"id":"xai/grok-3-mini","name":"Grok 3 Mini","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"release_date":"2025-04-11","last_updated":"2025-04-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":0.3,"output":0.5,"cache_read":0.075}},"xai/grok-4.1-fast-reasoning":{"id":"xai/grok-4.1-fast-reasoning","name":"Grok-4.1-Fast-Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast-reasoning"},"xai/grok-4-fast-non-reasoning":{"id":"xai/grok-4-fast-non-reasoning","name":"Grok-4-Fast-Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"status":"deprecated","cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"xai/grok-4.20-multi-agent":{"id":"xai/grok-4.20-multi-agent","name":"Grok-4.20-Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high"]}],"tool_call":true,"temperature":false,"release_date":"2026-03-13","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":0},"cost":{"input":1.2626,"output":2.5253,"cache_read":0.202}},"xai/grok-code-fast-1":{"id":"xai/grok-code-fast-1","name":"Grok Code Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.2,"output":1.5,"cache_read":0.02},"canonical_model_id":"xai/grok-code-fast-1"},"xai/grok-4":{"id":"xai/grok-4","name":"Grok-4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-07-10","last_updated":"2025-07-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.75}},"xai/grok-4.1-fast-non-reasoning":{"id":"xai/grok-4.1-fast-non-reasoning","name":"Grok-4.1-Fast-Non-Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"status":"deprecated","canonical_model_id":"xai/grok-4.1-fast"},"xai/grok-3":{"id":"xai/grok-3","name":"Grok 3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-04-11","last_updated":"2025-04-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.75}},"xai/grok-4-fast-reasoning":{"id":"xai/grok-4-fast-reasoning","name":"Grok-4-Fast-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":128000},"status":"deprecated","cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-reasoning"},"novita/glm-4.6":{"id":"novita/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"cost":{"input":0.6061,"output":2.2222,"cache_read":0.1111},"canonical_model_id":"zhipuai/glm-4.6"},"novita/glm-4.7-flash":{"id":"novita/glm-4.7-flash","name":"glm-4.7-flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":65500},"cost":{"input":0.0707,"output":0.404},"canonical_model_id":"zhipuai/glm-4.7-flash"},"novita/glm-4.7":{"id":"novita/glm-4.7","name":"glm-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"status":"deprecated","canonical_model_id":"zhipuai/glm-4.7"},"novita/kimi-k2.5":{"id":"novita/kimi-k2.5","name":"Kimi-K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":262144},"status":"deprecated","cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"novita/minimax-m2.1":{"id":"novita/minimax-m2.1","name":"minimax-m2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"cost":{"input":0.303,"output":1.2121,"cache_read":0.1515},"canonical_model_id":"minimax/MiniMax-M2.1"},"novita/glm-4.7-n":{"id":"novita/glm-4.7-n","name":"glm-4.7-n","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"canonical_model_id":"zhipuai/glm-4.7"},"novita/kimi-k2.6":{"id":"novita/kimi-k2.6","name":"Kimi-K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.9596,"output":4.0404,"cache_read":0.1616},"canonical_model_id":"moonshotai/kimi-k2.6"},"novita/glm-5":{"id":"novita/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"cost":{"input":1,"output":3.2,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"novita/glm-4.6v":{"id":"novita/glm-4.6v","name":"glm-4.6v","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":32768},"canonical_model_id":"zhipuai/glm-4.6v"},"novita/kimi-k2-thinking":{"id":"novita/kimi-k2-thinking","name":"kimi-k2-thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":0},"status":"deprecated","canonical_model_id":"moonshotai/kimi-k2-thinking"},"novita/deepseek-v3.2":{"id":"novita/deepseek-v3.2","name":"DeepSeek-V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":0},"cost":{"input":0.2626,"output":0.3838,"cache_read":0.1313},"canonical_model_id":"deepseek/deepseek-v3.2"},"lumalabs/ray2":{"id":"lumalabs/ray2","type":"video","name":"Ray2","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"ray","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-02-20","last_updated":"2025-02-20","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":5000,"output":0}},"stabilityai/stablediffusionxl":{"id":"stabilityai/stablediffusionxl","type":"image","name":"StableDiffusionXL","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"stable-diffusion","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2023-07-09","last_updated":"2023-07-09","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":200,"output":0}},"elevenlabs/elevenlabs-v2.5-turbo":{"id":"elevenlabs/elevenlabs-v2.5-turbo","type":"speech","name":"ElevenLabs-v2.5-Turbo","description":"Low-latency multilingual text-to-speech model, superseded by Flash v2.5","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-07-19","last_updated":"2024-07-19","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"elevenlabs/eleven_turbo_v2_5"},"elevenlabs/elevenlabs-music":{"id":"elevenlabs/elevenlabs-music","name":"ElevenLabs-Music","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-08-29","last_updated":"2025-08-29","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":2000,"output":0}},"elevenlabs/elevenlabs-v3":{"id":"elevenlabs/elevenlabs-v3","type":"speech","name":"ElevenLabs-v3","description":"Expressive text-to-speech model with inline audio tags across 70+ languages","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-08-20","last_updated":"2025-08-20","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"elevenlabs/eleven_v3"},"cerebras/llama-3.3-70b-cs":{"id":"cerebras/llama-3.3-70b-cs","name":"llama-3.3-70b-cs","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"status":"deprecated","canonical_model_id":"meta/llama-3.3-70b-instruct"},"cerebras/qwen3-32b-cs":{"id":"cerebras/qwen3-32b-cs","name":"qwen3-32b-cs","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"status":"deprecated","canonical_model_id":"alibaba/qwen3-32b"},"cerebras/llama-3.1-8b-cs":{"id":"cerebras/llama-3.1-8b-cs","name":"Llama-3.1-8B-CS","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":0},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"cerebras/gpt-oss-120b-cs":{"id":"cerebras/gpt-oss-120b-cs","name":"GPT-OSS-120B-CS","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":0},"cost":{"input":0.3535,"output":0.7576},"canonical_model_id":"openai/gpt-oss-120b"},"cerebras/qwen3-235b-2507-cs":{"id":"cerebras/qwen3-235b-2507-cs","name":"qwen3-235b-2507-cs","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":0,"output":0},"status":"deprecated","canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"runwayml/runway":{"id":"runwayml/runway","type":"video","name":"Runway","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"runway","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-10-11","last_updated":"2024-10-11","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":256,"output":0},"status":"deprecated"},"runwayml/runway-gen-4-turbo":{"id":"runwayml/runway-gen-4-turbo","type":"video","name":"Runway-Gen-4-Turbo","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"runway","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-05-09","last_updated":"2025-05-09","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":256,"output":0},"status":"deprecated"},"openai/gpt-image-1.5":{"id":"openai/gpt-image-1.5","type":"image","name":"gpt-image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":128000,"output":0},"canonical_model_id":"openai/gpt-image-1.5"},"openai/gpt-5.1-instant":{"id":"openai/gpt-5.1-instant","name":"GPT-5.1-Instant","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":14,"output":54},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1-mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.3636,"output":1.4545,"cache_read":0.0909},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1-nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.09,"output":0.36,"cache_read":0.022},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.2-instant":{"id":"openai/gpt-5.2-instant","name":"GPT-5.2-Instant","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":1.6,"output":13,"cache_read":0.16},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"openai/gpt-image-1-mini":{"id":"openai/gpt-image-1-mini","type":"image","name":"GPT-Image-1-Mini","description":"Cost-efficient GPT Image 1 variant for image generation and editing","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"canonical_model_id":"openai/gpt-image-1-mini"},"openai/gpt-4o-aug":{"id":"openai/gpt-4o-aug","name":"GPT-4o-Aug","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2.2727,"output":9.0909,"cache_read":1.1364},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1364,"output":9.0909,"cache_read":0.1136},"canonical_model_id":"openai/gpt-5.1"},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":140,"output":540},"canonical_model_id":"openai/o1-pro"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1-Codex-Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2727,"output":13.6364,"cache_read":0.2273},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-5.3-codex-spark":{"id":"openai/gpt-5.3-codex-spark","name":"GPT-5.3-Codex-Spark","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-spark","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":100000,"output":16384},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.3-codex-spark"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":1.8182,"output":7.2727,"cache_read":0.4545},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1-Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-4o-search":{"id":"openai/gpt-4o-search","name":"GPT-4o-Search","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-03-11","last_updated":"2025-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":2.2727,"output":9.0909}},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2273,"output":1.8182,"cache_read":0.0227},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2-Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":19.0909,"output":152.7273},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":922000,"output":128000},"cost":{"input":4.5455,"output":27.2727,"cache_read":0.4545},"canonical_model_id":"openai/gpt-5.5"},"openai/chatgpt-4o-latest":{"id":"openai/chatgpt-4o-latest","name":"ChatGPT-4o-Latest","description":"Legacy model retained for compatibility with older integrations","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-08-14","last_updated":"2024-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":4.5,"output":14}},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5-nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.0455,"output":0.3636,"cache_read":0.0045},"canonical_model_id":"openai/gpt-5-nano"},"openai/o3-deep-research":{"id":"openai/o3-deep-research","name":"o3-deep-research","description":"Research model for long-horizon investigation, synthesis, and analytical reports","family":"o","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":9,"output":36,"cache_read":2.2},"canonical_model_id":"openai/o3-deep-research"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4-Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.6818,"output":4.0909,"cache_read":0.0682},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5-Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":922000,"output":128000},"cost":{"input":27.2727,"output":163.6364},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.5909,"output":12.7273,"cache_read":0.1591},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":0.99,"output":4},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.5909,"output":12.7273,"cache_read":0.1591},"canonical_model_id":"openai/gpt-5.2"},"openai/dall-e-3":{"id":"openai/dall-e-3","type":"image","name":"DALL-E-3","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"dall-e","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":800,"output":0},"status":"deprecated"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4-Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":27.2727,"output":163.6364},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5-Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":13.6364,"output":109.0909},"canonical_model_id":"openai/gpt-5-pro"},"openai/gpt-3.5-turbo-raw":{"id":"openai/gpt-3.5-turbo-raw","name":"GPT-3.5-Turbo-Raw","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":4524,"output":2048},"status":"deprecated","cost":{"input":0.45,"output":1.4},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-image-1":{"id":"openai/gpt-image-1","type":"image","name":"GPT-Image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":128000,"output":0},"status":"deprecated","canonical_model_id":"openai/gpt-image-1"},"openai/gpt-5.3-instant":{"id":"openai/gpt-5.3-instant","name":"GPT-5.3-Instant","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":111616,"output":16384},"status":"deprecated","cost":{"input":1.6,"output":13,"cache_read":0.16},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"openai/gpt-5-chat":{"id":"openai/gpt-5-chat","name":"GPT-5-Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":16384},"status":"deprecated","cost":{"input":1.1,"output":9,"cache_read":0.11},"canonical_model_id":"openai/gpt-5-chat-latest"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":0.99,"output":4,"cache_read":0.25},"canonical_model_id":"openai/o4-mini"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":2048},"status":"deprecated","cost":{"input":0.45,"output":1.4},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4-Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.1818,"output":1.1364,"cache_read":0.0182},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1-Codex-Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.22,"output":1.8,"cache_read":0.022},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"canonical_model_id":"openai/gpt-4o"},"openai/sora-2-pro":{"id":"openai/sora-2-pro","type":"video","name":"Sora-2-Pro","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"sora","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"status":"deprecated"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.8182,"output":7.2727,"cache_read":0.4545},"canonical_model_id":"openai/o3"},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o-mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":124096,"output":4096},"cost":{"input":0.1364,"output":0.5455,"cache_read":0.0682},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-image-2":{"id":"openai/gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5.0505,"output":30.303,"cache_read":1.2626},"canonical_model_id":"openai/gpt-image-2"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2-Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.6,"output":13,"cache_read":0.16},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-3.5-turbo-instruct":{"id":"openai/gpt-3.5-turbo-instruct","name":"GPT-3.5-Turbo-Instruct","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2023-09-20","last_updated":"2023-09-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":3500,"output":1024},"status":"deprecated","cost":{"input":1.4,"output":1.8}},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.1364,"output":9.0909,"cache_read":0.1136},"canonical_model_id":"openai/gpt-5"},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":18.1818,"output":72.7273},"canonical_model_id":"openai/o3-pro"},"openai/gpt-4o-mini-search":{"id":"openai/gpt-4o-mini-search","name":"GPT-4o-mini-Search","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-03-11","last_updated":"2025-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.14,"output":0.54}},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4-Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":9,"output":27},"canonical_model_id":"openai/gpt-4-turbo"},"openai/o4-mini-deep-research":{"id":"openai/o4-mini-deep-research","name":"o4-mini-deep-research","description":"Research model for long-horizon investigation, synthesis, and analytical reports","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":1.8,"output":7.2,"cache_read":0.45},"canonical_model_id":"openai/o4-mini-deep-research"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.1,"output":9},"canonical_model_id":"openai/gpt-5-codex"},"openai/sora-2":{"id":"openai/sora-2","type":"video","name":"Sora-2","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"sora","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":0,"output":0},"status":"deprecated"},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"o3-mini-high","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"status":"deprecated","cost":{"input":0.99,"output":4},"canonical_model_id":"openai/o3-mini"},"poetools/claude-code":{"id":"poetools/claude-code","name":"claude-code","description":"Claude model for careful reasoning, writing, coding, and tool use","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]},{"type":"budget_tokens"}],"tool_call":true,"temperature":false,"release_date":"2025-11-27","last_updated":"2025-11-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"anthropic/claude-sonnet-3.5-june":{"id":"anthropic/claude-sonnet-3.5-june","name":"Claude-Sonnet-3.5-June","description":"Legacy model retained for compatibility with older integrations","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-11-18","last_updated":"2024-11-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2}},"anthropic/claude-sonnet-3.7":{"id":"anthropic/claude-sonnet-3.7","name":"Claude-Sonnet-3.7","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-10-31","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":128000},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2},"canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"anthropic/claude-sonnet-4":{"id":"anthropic/claude-sonnet-4","name":"Claude-Sonnet-4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":64000},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude-Opus-4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":128000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude-Opus-4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":31999}],"tool_call":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":32000},"status":"deprecated","cost":{"input":13,"output":64,"cache_read":1.3,"cache_write":16},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude-Haiku-4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":63999}],"tool_call":true,"temperature":false,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":192000,"output":64000},"cost":{"input":0.8586,"output":4.2929,"cache_read":0.0859,"cache_write":1.0732},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude-Opus-4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":0,"max":63999}],"tool_call":true,"temperature":false,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":196608,"output":64000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude-Sonnet-4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":0,"max":31999}],"tool_call":true,"temperature":false,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":32768},"cost":{"input":2.5758,"output":12.8788,"cache_read":0.2576,"cache_write":3.2197},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4":{"id":"anthropic/claude-opus-4","name":"Claude-Opus-4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":192512,"output":28672},"status":"deprecated","cost":{"input":13,"output":64,"cache_read":1.3,"cache_write":16},"canonical_model_id":"anthropic/claude-opus-4-0"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude-Opus-4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":128000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-haiku-3.5":{"id":"anthropic/claude-haiku-3.5","name":"Claude-Haiku-3.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-07-31","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":0.68,"output":3.4,"cache_read":0.068,"cache_write":0.85},"canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"anthropic/claude-haiku-3":{"id":"anthropic/claude-haiku-3","name":"Claude-Haiku-3","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2023-08-31","release_date":"2024-03-13","last_updated":"2024-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":0.21,"output":1.1,"cache_read":0.021,"cache_write":0.26},"canonical_model_id":"anthropic/claude-3-haiku-20240307"},"anthropic/claude-sonnet-3.5":{"id":"anthropic/claude-sonnet-3.5","name":"Claude-Sonnet-3.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-04-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":189096,"output":8192},"status":"deprecated","cost":{"input":2.6,"output":13,"cache_read":0.26,"cache_write":3.2},"canonical_model_id":"anthropic/claude-3-5-sonnet-20241022"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude-Opus-4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":128000},"cost":{"input":4.2929,"output":21.4646,"cache_read":0.4293,"cache_write":5.3662},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude-Sonnet-4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":983040,"output":128000},"cost":{"input":2.5758,"output":12.8788,"cache_read":0.2576,"cache_write":3.2197},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"trytako/tako":{"id":"trytako/tako","name":"Tako","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"tako","attachment":false,"reasoning":false,"tool_call":true,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":0}},"fireworks-ai/kimi-k2.5-fw":{"id":"fireworks-ai/kimi-k2.5-fw","name":"Kimi-K2.5-FW","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":245760,"output":16384},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"moonshotai/kimi-k2.5"},"empiriolabs/deepseek-v4-flash-el":{"id":"empiriolabs/deepseek-v4-flash-el","name":"DeepSeek-V4-Flash-EL","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":0.14,"output":0.28},"canonical_model_id":"deepseek/deepseek-v4-flash"},"empiriolabs/deepseek-v4-pro-el":{"id":"empiriolabs/deepseek-v4-pro-el","name":"DeepSeek-V4-Pro-EL","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":1.67,"output":3.33},"canonical_model_id":"deepseek/deepseek-v4-pro"}}},"cortecs":{"id":"cortecs","env":["CORTECS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.cortecs.ai/v1","name":"Cortecs","doc":"https://api.cortecs.ai/v1/models","models":{"gemma-3-27b-it":{"id":"gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Gemma 3 is a family of lightweight, multimodal models from Google, supporting text and image inputs, multilingual capabilities, and a 131K context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":110000},"cost":{"input":0.099,"output":0.299},"canonical_model_id":"google/gemma-3-27b-it"},"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.111,"output":0.167},"canonical_model_id":"alibaba/qwen3.5-9b"},"voxtral-small-2507":{"id":"voxtral-small-2507","name":"voxtral-small-2507","description":"Voxtral Small is a multimodal model with audio input, combining advanced speech capabilities with strong text performance for transcription, translation, and audio understanding.","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.123,"output":0.368,"cache_read":0.012},"canonical_model_id":"mistral/voxtral-small-24b-2507"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.434,"output":1.704,"cache_read":0.134},"canonical_model_id":"openai/gpt-4.1-mini"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.656,"output":3.3,"cache_read":0.18},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":203000,"output":203000},"cost":{"input":0.08,"output":0.478},"canonical_model_id":"zhipuai/glm-4.7-flash"},"gemma-4-31b-it":{"id":"gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.223,"output":0.39},"canonical_model_id":"google/gemma-4-31b-it"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.111,"output":0.434,"cache_read":0.056},"canonical_model_id":"openai/gpt-4.1-nano"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":2.25,"output":11.249,"cache_read":0.113},"canonical_model_id":"google/gemini-4-argon"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.089,"output":0.446,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-120b"},"claude-opus4-5":{"id":"claude-opus4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5.313,"output":26.568,"cache_read":0.531,"cache_write":6.645},"canonical_model_id":"anthropic/claude-opus-4-5"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.724,"output":0.724},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":10.96,"cache_read":0.156},"canonical_model_id":"openai/gpt-5.1"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1000000},"cost":{"input":0.9,"output":3.24,"cache_read":0.189},"canonical_model_id":"zhipuai/glm-5.2"},"gpt-oss-safeguard-120b":{"id":"gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.179,"output":0.697},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"nova-micro-v1":{"id":"nova-micro-v1","name":"Nova Micro 1.0","description":"Nova Micro is a multilingual text-to-text foundation model with strong reasoning capabilities and broad language coverage across 200+ languages.","family":"nova-micro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000},"cost":{"input":0.04,"output":0.159},"canonical_model_id":"amazon/nova-micro-v1"},"mistral-large-4-0":{"id":"mistral-large-4-0","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":0.765,"output":2.351,"cache_read":0.079},"canonical_model_id":"mistral/mistral-large-4"},"ministral-14b-2512":{"id":"ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Ministral 3 14B is a frontier-level 14B multimodal model optimized for local deployment, delivering state-of-the-art text and vision reasoning with a 256K context window and strong agentic capabilities.","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.24,"output":0.24,"cache_read":0.022},"canonical_model_id":"mistral/ministral-14b-2512"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.649,"output":9.899,"cache_read":0.165,"cache_write":1},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"cost":{"input":2.898,"output":15.453,"cache_read":0.242},"canonical_model_id":"openai/gpt-5.4"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":12.272,"output":61.359,"cache_read":1.228,"cache_write":15.34},"canonical_model_id":"openai/gpt-6-astra"},"devstral-2512":{"id":"devstral-2512","name":"Devstral 2","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.478,"output":2.392,"cache_read":0.045},"canonical_model_id":"mistral/devstral-2512"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.825,"output":4.125,"cache_read":0.082,"cache_write":0.084},"canonical_model_id":"google/gemini-3.8-flash"},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.179,"output":0.697,"cache_read":0.014},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":393216},"cost":{"input":0.201,"output":0.6,"cache_read":0.004},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"qwen2.5-vl-72b-instruct":{"id":"qwen2.5-vl-72b-instruct","name":"qwen2.5-vl-72b-instruct","description":"Qwen2.5-VL is a powerful vision-language model with advanced capabilities in visual understanding, long video reasoning, and structured output generation.","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":false,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":1.014,"output":1.014},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2.192,"output":8.769,"cache_read":0.546},"canonical_model_id":"openai/gpt-4.1"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.279,"output":1.114,"cache_read":0.111},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.7,"output":3.4,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v4-pro"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.038},"canonical_model_id":"google/gemini-3.7-flash"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.045,"output":0.167},"canonical_model_id":"openai/gpt-oss-20b"},"nvidia-nemotron-3-nano-30b-a3b":{"id":"nvidia-nemotron-3-nano-30b-a3b","name":"nvidia-nemotron-3-nano-30b-a3b","description":"Nemotron-Nano-3-30B-A3B is a compact Mixture-of-Experts model optimized for efficient reasoning, chat, and coding, with strong multilingual support and long-context RAG and agent workflows.","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.06,"output":0.24},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nova-2-lite":{"id":"nova-2-lite","name":"Nova 2 Lite","description":"Nova 2 Lite is an advanced multimodal reasoning model that combines efficiency and performance, delivering reliable AI for agentic workflows and enterprise applications.","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535},"cost":{"input":0.373,"output":3.144},"canonical_model_id":"amazon/nova-2-lite"},"glm-5.1":{"id":"glm-5.1","name":"GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":202752},"cost":{"input":1.384,"output":4.348,"cache_read":0.346},"canonical_model_id":"zhipuai/glm-5.1"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.2,"output":11,"cache_read":0.219,"cache_write":2.749},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-opus4-6":{"id":"claude-opus4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.313,"output":26.561,"cache_read":0.531,"cache_write":6.645},"canonical_model_id":"anthropic/claude-opus-4-6"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.279,"output":2.192,"cache_read":0.056},"canonical_model_id":"openai/gpt-5-mini"},"qwen3.6-27b":{"id":"qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.446,"output":3.008},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen3-30b-a3b-instruct-2507":{"id":"qwen3-30b-a3b-instruct-2507","name":"qwen3-30b-a3b-instruct-2507","description":"Qwen3-30B-A3B-Instruct-2507 is an advanced Mixture-of-Experts model optimized for reasoning, coding, and multilingual instruction following.","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.099,"output":0.299},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-flash-next":{"id":"qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":64000},"cost":{"input":0.201,"output":0.5,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-flash-next"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":2.5,"output":6,"cache_read":0.625},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"codestral-2508":{"id":"codestral-2508","name":"Codestral 25.08","description":"Mistral coding model for code completion, generation, and developer workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":0.368,"output":1.103,"cache_read":0.037},"canonical_model_id":"mistral/codestral-2508"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.5,"output":27.498,"cache_read":0.55,"cache_write":6.874},"canonical_model_id":"anthropic/claude-opus-5"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.299,"output":2.491,"cache_read":0.029,"cache_write":0.097},"canonical_model_id":"google/gemini-2.5-flash"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0.996,"output":4.982,"cache_read":0.099,"cache_write":1.186},"canonical_model_id":"anthropic/claude-haiku-4-5"},"qwen3guard-gen-8b":{"id":"qwen3guard-gen-8b","name":"qwen3guard-gen-8b","description":"Qwen3Guard-Gen-8B is a large-scale multilingual safety moderation model designed for high-accuracy prompt and response classification.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0,"output":0}},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3 Coder Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.167,"output":0.891},"canonical_model_id":"alibaba/qwen3-coder-next"},"minimax-m2.1":{"id":"minimax-m2.1","name":"MiniMax-M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"output":196000},"cost":{"input":0.359,"output":1.435},"canonical_model_id":"minimax/MiniMax-M2.1"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.12,"output":0.6,"cache_read":0.012,"cache_write":0.15},"canonical_model_id":"openai/gpt-6-luna"},"mistral-7b-instruct-v0.2":{"id":"mistral-7b-instruct-v0.2","name":"mistral-7b-instruct-v0.2","description":"Mistral 7B Instruct is a compact, 7B parameter model optimized for fast and efficient text and code generation with a 32K token context window.","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2025-05-26","last_updated":"2025-05-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":0.159,"output":0.219}},"mistral-large-2402":{"id":"mistral-large-2402","name":"mistral-large-2402","description":"Mistral Large (24.02) is Mistral AI’s most advanced language model, built for complex multilingual reasoning, code generation, and deep text understanding.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-05-26","last_updated":"2025-05-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":4.284,"output":12.952}},"hermes-4-405b":{"id":"hermes-4-405b","name":"Hermes 4 405B","description":"Hermes 4 405B is a frontier hybrid-mode reasoning model built on Llama 3.1, optimized for advanced logic, math, coding, and structured output generation.","family":"hermes","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.996,"output":2.989},"canonical_model_id":"nousresearch/hermes-4-405b"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.06,"output":0.439,"cache_read":0.019},"canonical_model_id":"openai/gpt-5-nano"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":11.999,"cache_read":0.24,"cache_write":3},"canonical_model_id":"openai/gpt-6-sol"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":1.495,"output":9.964,"cache_read":0.242,"cache_write":0.434},"canonical_model_id":"google/gemini-2.5-pro"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":3,"output":14.999},"canonical_model_id":"moonshotai/kimi-k3"},"qwen3-235b-a22b-instruct-2507":{"id":"qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.199,"output":0.598,"cache_read":0.018},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"mixtral-8x7B-instruct-v0.1":{"id":"mixtral-8x7B-instruct-v0.1","name":"Mixtral 8x7B Instruct v0.1","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mistral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.488,"output":0.758},"canonical_model_id":"mistral/mixtral-8x7b-instruct-v0.1"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.475,"output":2.97,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.6"},"nova-pro-v1":{"id":"nova-pro-v1","name":"Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.918,"output":3.671},"canonical_model_id":"amazon/nova-pro-v1"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":64000},"cost":{"input":2,"output":3.999,"cache_read":0.5},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"claude-opus-5.5":{"id":"claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4.399,"output":21.998,"cache_read":0.219,"cache_write":5.5},"canonical_model_id":"anthropic/claude-opus-5-5"},"ministral-8b-2512":{"id":"ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Ministral 3 8B is a balanced, efficient multimodal model offering strong text and vision capabilities, optimized for edge and local deployment.","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.179,"output":0.179,"cache_read":0.017},"canonical_model_id":"mistral/ministral-8b-2512"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.082,"output":0.413,"cache_read":0.06},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.2,"output":13.199,"cache_read":0.219,"cache_write":2.749},"canonical_model_id":"openai/gpt-5.6-terra"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.668,"output":4.01},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.4,"output":11.999,"cache_read":0.12,"cache_write":3},"canonical_model_id":"openai/gpt-6.1-sol"},"pixtral-large-2502":{"id":"pixtral-large-2502","name":"Pixtral Large (25.02)","description":"Pixtral Large (25.02) is a 124B open-weight multimodal model built on Mistral Large 2, offering advanced image understanding and strong performance across text and code tasks.","family":"pixtral","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"cost":{"input":1.993,"output":5.978},"canonical_model_id":"mistral/pixtral-large-2502"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.511,"output":2.045,"cache_read":0.154},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.179,"output":0.697},"canonical_model_id":"alibaba/qwen3-32b"},"kolibri-1":{"id":"kolibri-1","name":"kolibri-1","description":"Kolibri-1 is Aleph Alpha’s open-weight English-German MoE language model, designed for enterprise use with efficient inference and advanced reasoning.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-22","last_updated":"2026-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":81920},"cost":{"input":0.167,"output":0.613,"cache_read":0.084}},"nova-lite-v1":{"id":"nova-lite-v1","name":"Nova Lite 1.0","description":"Nova Lite is a fast, low-cost multimodal foundation model capable of reasoning over text, images, and video in 200+ languages.","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000},"cost":{"input":0.069,"output":0.275},"canonical_model_id":"amazon/nova-lite-v1"},"mistral-nemo-instruct-2407":{"id":"mistral-nemo-instruct-2407","name":"mistral-nemo-instruct-2407","description":"A 12B parameter, instruct-tuned language model by Mistral AI and NVIDIA, designed for advanced instruction following, multi-turn conversations, and generating text and code across multiple languages.","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":118000,"output":118000},"cost":{"input":0.145,"output":0.145,"cache_read":0.014},"canonical_model_id":"mistral/mistral-nemo"},"mistral-small-2503":{"id":"mistral-small-2503","name":"mistral-small-2503","description":"Combines advanced text and vision capabilities with 24 billion parameters, supporting multilingual tasks and long contexts up to 131k tokens, making it versatile for various applications without sacrificing performance.","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.111,"output":0.334},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.038},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":2.659,"output":10.635,"cache_read":1.33},"canonical_model_id":"openai/gpt-4o"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"output":196000},"cost":{"input":0.348,"output":1.39,"cache_read":0.075},"canonical_model_id":"minimax/MiniMax-M2.5"},"claude-4-5-sonnet":{"id":"claude-4-5-sonnet","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":2.989,"output":14.945,"cache_read":0.326,"cache_write":4.078},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"claude-opus4-8":{"id":"claude-opus4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.437,"output":27.186,"cache_read":0.544,"cache_write":6.797},"canonical_model_id":"anthropic/claude-opus-4-8"},"claude-haiku-5.5":{"id":"claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.113,"output":0.563,"cache_read":0.011,"cache_write":0.14},"canonical_model_id":"anthropic/claude-haiku-5-5"},"mistral-small-2603":{"id":"mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.156,"output":0.625,"cache_read":0.016},"canonical_model_id":"mistral/mistral-small-2603"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":32768},"cost":{"input":0.167,"output":0.557},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"nemotron-nano-v2-12b":{"id":"nemotron-nano-v2-12b","name":"nemotron-nano-v2-12b","description":"NVIDIA Nemotron Nano v2 12B is a 12-billion-parameter multimodal reasoning model designed for advanced video understanding, document intelligence, and visual reasoning, built with a hybrid Transformer-Mamba architecture for high efficiency and low latency.","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.24,"output":0.707},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.219,"output":1.32,"cache_read":0.022,"cache_write":0.275},"canonical_model_id":"openai/gpt-5.6-luna"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.5,"output":1.499,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v3.2"},"gemma-4-26b-a4b-it":{"id":"gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":81920},"cost":{"input":0.111,"output":0.557},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16000},"cost":{"input":0.159,"output":0.638,"cache_read":0.081},"canonical_model_id":"openai/gpt-4o-mini"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.1,"output":0.35,"cache_read":0.018},"canonical_model_id":"zhipuai/glm-5.3-flash"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.33,"output":2.749,"cache_read":0.033,"cache_write":0.082},"canonical_model_id":"google/gemini-3.5-flash-lite"},"claude-sonnet-5.5":{"id":"claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2.25,"output":11.249,"cache_read":0.225,"cache_write":2.813},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.375,"output":10.96,"cache_read":0.156},"canonical_model_id":"openai/gpt-5"},"mistral-large-2512":{"id":"mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.613,"output":1.838,"cache_read":0.061},"canonical_model_id":"mistral/mistral-large-2512"},"claude-opus4-7":{"id":"claude-opus4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5.437,"output":27.186,"cache_read":0.544,"cache_write":6.797},"canonical_model_id":"anthropic/claude-opus-4-7"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.272,"output":1.631,"cache_read":0.025,"cache_write":0.082},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4.399,"output":21.998,"cache_read":0.44,"cache_write":5.5},"canonical_model_id":"openai/gpt-5.6-sol"},"apertus-70b":{"id":"apertus-70b","name":"Apertus 70B","description":"Apertus 70B is an open, multilingual language model designed for research, long-context reasoning, and sovereignty-focused AI systems.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"cost":{"input":1.393,"output":2.228},"canonical_model_id":"swiss-ai/apertus-70b"},"mistral-medium-3.5":{"id":"mistral-medium-3.5","name":"mistral-medium-3.5","description":"Mistral Medium 3.5 is a frontier multimodal 128B model combining reasoning, coding, and instruction-following with strong agentic performance and efficient deployment.","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1.532,"output":7.843,"cache_read":0.154},"canonical_model_id":"mistral/mistral-medium-3-5"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":500000},"cost":{"input":0.1,"output":0.4,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"mistral-small-3.2-24b-instruct-2506":{"id":"mistral-small-3.2-24b-instruct-2506","name":"mistral-small-3.2-24b-instruct-2506","description":"Mistral-Small-3.2-24B-Instruct-2506 is a 24B parameter instruction-tuned model with enhanced long-context support (128k) and state-of-the-art vision understanding.","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.1,"output":0.312},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"ministral-3b-2512":{"id":"ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Ministral 3 3B is a compact, efficient multimodal model with strong language, vision capabilities, and ideal for custom fine-tuning.","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.123,"output":0.123,"cache_read":0.012},"canonical_model_id":"mistral/ministral-3b-2512"},"qwen3-vl-235b-a22b":{"id":"qwen3-vl-235b-a22b","name":"qwen3-vl-235b-a22b","description":"Qwen3 VL 235B A22B is a 235B-parameter MoE vision-language flagship model (≈22B active) designed for frontier-level multimodal understanding across text, images, documents, and long videos.","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.617,"output":3.119,"cache_read":0.052},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"minimax-m2":{"id":"minimax-m2","name":"MiniMax-M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":400000,"output":196000},"cost":{"input":0.349,"output":1.405},"canonical_model_id":"minimax/MiniMax-M2"},"qwen3guard-gen-0.6b":{"id":"qwen3guard-gen-0.6b","name":"qwen3guard-gen-0.6b","description":"Qwen3Guard-Gen-0.6B is a lightweight multilingual safety moderation model that classifies prompts and responses into safe, controversial, or unsafe categories.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-04","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0,"output":0}},"claude-4-6-sonnet":{"id":"claude-4-6-sonnet","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":3.196,"output":15.94,"cache_read":0.32,"cache_write":3.999},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"minicpm-v-4.5":{"id":"minicpm-v-4.5","name":"minicpm-v-4.5","description":"MiniCPM-V 4.5 is a compact, high-performance vision-language model excelling in video understanding, OCR, and multimodal reasoning with efficient deployment.","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.651,"output":1.097}}}},"zenmux":{"id":"zenmux","env":["ZENMUX_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://zenmux.ai/api/v1","name":"ZenMux","doc":"https://docs.zenmux.ai","models":{"volcengine/doubao-seed-2.0-lite":{"id":"volcengine/doubao-seed-2.0-lite","name":"Doubao-Seed-2.0-lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02-14","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.09,"output":0.51,"cache_read":0.02,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"volcengine/doubao-seed-2.0-mini":{"id":"volcengine/doubao-seed-2.0-mini","name":"Doubao-Seed-2.0-mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02-14","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.03,"output":0.28,"cache_read":0.01,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"volcengine/doubao-seed-2.0-code":{"id":"volcengine/doubao-seed-2.0-code","name":"Doubao Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":0.9,"output":4.48},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"volcengine/doubao-seed-code":{"id":"volcengine/doubao-seed-code","name":"Doubao-Seed-Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-11-11","last_updated":"2025-11-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.17,"output":1.12,"cache_read":0.03}},"volcengine/doubao-seed-1.8":{"id":"volcengine/doubao-seed-1.8","name":"Doubao-Seed-1.8","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.11,"output":0.28,"cache_read":0.02,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-1-8"},"volcengine/doubao-seed-2.0-pro":{"id":"volcengine/doubao-seed-2.0-pro","name":"Doubao-Seed-2.0-pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02-14","release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.45,"output":2.24,"cache_read":0.09,"cache_write":0.0024},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.726,"output":3.1946,"cache_read":0.1743,"tiers":[{"input":1.0165,"output":3.7754,"cache_read":0.2614,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.2911,"output":1.1645,"cache_read":0.0582,"tiers":[{"input":0.5823,"output":2.3291,"cache_read":0.1165,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.5-air":{"id":"z-ai/glm-4.5-air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.1165,"output":0.2911,"cache_read":0.0233,"tiers":[{"input":0.1747,"output":1.1645,"cache_read":0.0349,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.2911,"output":1.1645,"cache_read":0.0582,"tiers":[{"input":0.5823,"output":2.3291,"cache_read":0.1165,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.8781,"output":3.5126,"cache_read":0.1903,"tiers":[{"input":1.1709,"output":4.098,"cache_read":0.2927,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":96000},"cost":{"input":0.2911,"output":1.1645,"cache_read":0.0582,"tiers":[{"input":0.5823,"output":2.3291,"cache_read":0.1165,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-4.7-flash-free":{"id":"z-ai/glm-4.7-flash-free","name":"GLM 4.7 Flash (Free)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-5.3-flashx":{"id":"z-ai/glm-5.3-flashx","name":"GLM 5.3 FlashX","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.375,"output":1.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-4.6v-flash":{"id":"z-ai/glm-4.6v-flash","name":"GLM 4.6V FlashX","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.0218,"output":0.2184,"cache_read":0.0044,"tiers":[{"input":0.0437,"output":0.4367,"cache_read":0.0044,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.58,"output":2.6,"cache_read":0.14,"tiers":[{"input":0.87,"output":3.18,"cache_read":0.22,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.7-flashx":{"id":"z-ai/glm-4.7-flashx","name":"GLM 4.7 FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.0728,"output":0.4367,"cache_read":0.0146},"canonical_model_id":"zhipuai/glm-4.7-flashx"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM 4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0.1456,"output":0.4367,"cache_read":0.0291,"tiers":[{"input":0.2911,"output":0.8734,"cache_read":0.0582,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-4.6v-flash-free":{"id":"z-ai/glm-4.6v-flash-free","name":"GLM 4.6V Flash (Free)","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"tiers":[{"input":0,"output":0,"cache_read":0,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"z-ai/glm-image":{"id":"z-ai/glm-image","type":"image","name":"GLM-Image","description":"GLM-Image is an image generation model adopts a hybrid autoregressive + diffusion decoder architecture. In general image generation quality, GLM‑Image aligns with mainstream latent diffusion approaches, but it shows significant advantages in text-rendering and knowledge‑intensive generation scenarios. It performs especially well in tasks requiring precise semantic understanding and complex information expression, while maintaining strong capabilities in high‑fidelity and fine‑grained detail generation. In addition to text‑to‑image generation, GLM‑Image also supports a rich set of image‑to‑image tasks including image editing, style transfer, identity‑preserving generation, and multi‑subject consistency.","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-14","last_updated":"2026-01-14","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":10240,"output":0},"provider":{"api":"https://zenmux.ai/api/vertex-ai"},"canonical_model_id":"zhipuai/glm-image"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":0.73,"output":3.19,"cache_read":0.174,"tiers":[{"input":1.02,"output":3.77,"cache_read":0.261,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0.075,"output":0.25,"cache_read":0.015},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":0.5,"output":3,"cache_read":0.05,"cache_write":1,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":1.25,"output":10,"cache_read":0.13,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":2,"cache_read":0.08,"tiers":[{"input":0.8,"output":4,"cache_read":0.16,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.8,"output":4,"cache_read":0.16}},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1,"output":3,"cache_read":0.2,"tiers":[{"input":2,"output":6,"cache_read":0.4,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.4}},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"deepseek/deepseek-v3.2-exp":{"id":"deepseek/deepseek-v3.2-exp","name":"DeepSeek-V3.2-Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":163000,"output":64000},"status":"deprecated","cost":{"input":0.22,"output":0.33}},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.075,"output":0.3,"cache_read":0.0015},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.66,"output":1.98,"cache_read":0.022},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":0.293,"output":0.4395,"cache_read":0.0293},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.007},"canonical_model_id":"deepseek/deepseek-v4-flash"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262140,"output":262140},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code HighSpeed","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.9,"output":8,"cache_read":0.38},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"baidu/ernie-5.0-thinking-preview":{"id":"baidu/ernie-5.0-thinking-preview","name":"ERNIE 5.0","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-01-22","last_updated":"2026-01-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":0.84,"output":3.37,"tiers":[{"input":1.41,"output":5.62,"tier":{"type":"context","size":32000}}]}},"inclusionai/ring-2.6-1t":{"id":"inclusionai/ring-2.6-1t","name":"inclusionAI: Ring-2.6-1T","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-12-31","release_date":"2026-05-07","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":65000},"cost":{"input":0.1318155,"output":1.0984625,"cache_read":0.0263631}},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2-Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["image","text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/chat-latest":{"id":"openai/chat-latest","name":"Chat Latest (GPT-5.5 Instant)","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-05-05","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5-instant"},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1-Codex-Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":100000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://zenmux.ai/api/v1"},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"minimax/minimax-m2.7-highspeed":{"id":"minimax/minimax-m2.7-highspeed","name":"MiniMax M2.7 highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131070},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.611,"output":2.4439,"cache_read":0.0611,"cache_write":0.3819},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m2.5-lightning":{"id":"minimax/minimax-m2.5-lightning","name":"MiniMax M2.5 highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.6,"output":2.4,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5-highspeed"},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":512000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"tiers":[{"input":0.6,"output":2.4,"cache_read":0.12,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.6,"output":2.4,"cache_read":0.12}},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204000,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131070},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://zenmux.ai/api/anthropic/v1"},"cost":{"input":0.3,"output":1.2,"cache_read":0.03,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":0.24,"cache_write":1.5}},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen3.6-Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625,"tiers":[{"input":2,"output":6,"cache_read":0.4,"cache_write":2.5,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":0.4,"cache_write":2.5}},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-01-01","release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1020000,"output":1020000},"cost":{"input":0.1,"output":0.4,"cache_read":0.02,"cache_write":0.125},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.4,"output":2.4,"cache_read":0.08,"cache_write":0.5,"tiers":[{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.5,"output":3,"cache_read":0.1,"cache_write":0.625}},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3-Max-Thinking","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3-coder-plus":{"id":"qwen/qwen3-coder-plus","name":"Qwen3-Coder-Plus","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"alibaba/qwen3-coder-plus"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":1.25,"output":3.75,"cache_read":0.25,"cache_write":1.5625},"canonical_model_id":"alibaba/qwen3.7-max"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-voice-tts-1.0":{"id":"x-ai/grok-voice-tts-1.0","type":"speech","name":"Grok Voice TTS 1.0","description":"Convert text into spoken audio with a single API call. The API supports a rich set of expressive voices, inline speech tags for fine-grained delivery control, and output formats from high-fidelity MP3 to telephony-optimized μ-law.","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":15000,"output":15000},"canonical_model_id":"xai/grok-tts"},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":1000000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-4.2-fast-non-reasoning":{"id":"x-ai/grok-4.2-fast-non-reasoning","name":"Grok 4.2 Fast Non Reasoning","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":4,"output":12,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"xai/grok-4.20-0309-non-reasoning"},"x-ai/grok-voice-stt-1.0":{"id":"x-ai/grok-voice-stt-1.0","type":"transcription","name":"Grok Voice STT 1.0","description":"Speech-to-text model with word-level timestamps, speaker diarization, multichannel audio, and streaming","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":15000,"output":15000},"canonical_model_id":"xai/grok-voice-transcribe-2.0"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-imagine-image-2.0":{"id":"x-ai/grok-imagine-image-2.0","type":"image","name":"Grok Imagine Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":66000,"output":0},"canonical_model_id":"xai/grok-imagine-image-2.0"},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.5"},"x-ai/grok-4.2-fast":{"id":"x-ai/grok-4.2-fast","name":"Grok 4.2 Fast","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":2,"output":6,"cache_read":0.2,"tiers":[{"input":4,"output":12,"cache_read":0.2,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"xai/grok-4.20-0309-reasoning"},"bytedance/doubao-seed-2.0-lite":{"id":"bytedance/doubao-seed-2.0-lite","name":"Doubao-Seed-2.0-lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.09,"output":0.51,"cache_read":0.02,"tiers":[{"input":0.13,"output":0.76,"cache_read":0.03,"tier":{"type":"context","size":32000}},{"input":0.25,"output":1.51,"cache_read":0.05,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance/doubao-seed-2.0-mini":{"id":"bytedance/doubao-seed-2.0-mini","name":"Doubao-Seed-2.0-mini","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.03,"output":0.28,"cache_read":0.01,"tiers":[{"input":0.06,"output":0.56,"cache_read":0.02,"tier":{"type":"context","size":32000}},{"input":0.12,"output":1.12,"cache_read":0.03,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"bytedance/doubao-seed-2.0-code":{"id":"bytedance/doubao-seed-2.0-code","name":"Doubao Seed 2.0 Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.45,"output":2.24,"cache_read":0.09,"tiers":[{"input":0.67,"output":3.36,"cache_read":0.14,"tier":{"type":"context","size":32000}},{"input":1.34,"output":6.71,"cache_read":0.27,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"bytedance/doubao-seed-code":{"id":"bytedance/doubao-seed-code","name":"Doubao-Seed-Code","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-11-11","last_updated":"2025-11-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.17,"output":1.12,"cache_read":0.034,"tiers":[{"input":0.2,"output":1.69,"cache_read":0.034,"tier":{"type":"context","size":32000}},{"input":0.39,"output":2.25,"cache_read":0.034,"tier":{"type":"context","size":128000}}]}},"bytedance/doubao-seed-1.8":{"id":"bytedance/doubao-seed-1.8","name":"Doubao-Seed-1.8","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.11,"output":0.28,"cache_read":0.023,"tiers":[{"input":0.17,"output":2.27,"cache_read":0.023,"tier":{"type":"context","size":32000}},{"input":0.34,"output":3.41,"cache_read":0.023,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-1-8"},"bytedance/doubao-seed-2.0-pro":{"id":"bytedance/doubao-seed-2.0-pro","name":"Doubao-Seed-2.0-pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000},"cost":{"input":0.45,"output":2.24,"cache_read":0.09,"tiers":[{"input":0.67,"output":3.36,"cache_read":0.14,"tier":{"type":"context","size":32000}},{"input":1.34,"output":6.71,"cache_read":0.27,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"}}},"cloudferro-sherlock":{"id":"cloudferro-sherlock","env":["CLOUDFERRO_SHERLOCK_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api-sherlock.cloudferro.com/openai/v1/","name":"CloudFerro Sherlock","doc":"https://docs.sherlock.cloudferro.com/","models":{"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":196000,"input":180000,"output":16000},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.5"},"meta-llama/Llama-3.3-70B-Instruct":{"id":"meta-llama/Llama-3.3-70B-Instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10-09","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":70000,"output":70000},"cost":{"input":2.92,"output":2.92},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"OpenAI GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":2.92,"output":2.92},"canonical_model_id":"openai/gpt-oss-120b"},"speakleash/Bielik-11B-v3.0-Instruct":{"id":"speakleash/Bielik-11B-v3.0-Instruct","name":"Bielik 11B v3.0 Instruct","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.67,"output":0.67}},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.34,"output":4.48},"canonical_model_id":"zhipuai/glm-5.3"}}},"perplexity-agent":{"id":"perplexity-agent","env":["PERPLEXITY_API_KEY"],"npm":"@ai-sdk/openai","api":"https://api.perplexity.ai/v1","name":"Perplexity Agent","doc":"https://docs.perplexity.ai/docs/agent-api/models","models":{"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"tiers":[{"input":0.5,"output":3,"cache_read":0.05,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.5,"output":3,"cache_read":0.05}},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":7.5,"cache_read":0.15},"canonical_model_id":"google/gemini-3.6-flash"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32000},"status":"deprecated","cost":{"input":0.25,"output":2.5},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"cache_write":2,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"xai/grok-4-1-fast-non-reasoning":{"id":"xai/grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"status":"deprecated","cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.3,"tiers":[{"input":4,"output":12,"cache_read":0.6,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":0.6}},"canonical_model_id":"xai/grok-4.5"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.13,"output":0.26,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.25,"output":2.5,"cache_read":0.0625},"canonical_model_id":"perplexity/sonar"},"perplexity/nemotron-3.5-super-120b-a12b":{"id":"perplexity/nemotron-3.5-super-120b-a12b","name":"Nemotron 3.5 Super 120B A12B","description":"NVIDIA Nemotron 3.5 Super reasoning model (120B total, 12B active) served with text-only input","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.2,"output":0.5,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-super-120b-a12b"},"perplexity/kimi-k3":{"id":"perplexity/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"perplexity/deepseek-v4-pro-0813":{"id":"perplexity/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"perplexity/glm-5.3":{"id":"perplexity/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"perplexity/glm-5.3-flash":{"id":"perplexity/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"moonshot-ai/kimi-k2.7-code":{"id":"moonshot-ai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshot-ai/kimi-k3":{"id":"moonshot-ai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8}},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"}}},"llmtech":{"id":"llmtech","env":["LLMTECH_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.llmtech.eu/v1","name":"LLM Tech","doc":"https://llmtech.eu/models/qwen3.8-27b","models":{"nvidia/Qwen3.8-27B-NVFP4":{"id":"nvidia/Qwen3.8-27B-NVFP4","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.25,"output":2.09,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"}}},"nova":{"id":"nova","env":["NOVA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.nova.amazon.com/v1","name":"Nova","doc":"https://nova.amazon.com/dev/documentation","models":{"nova-2-pro-v1":{"id":"nova-2-pro-v1","name":"Nova 2 Pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-12-03","last_updated":"2026-01-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"reasoning":0}},"nova-2-lite-v1":{"id":"nova-2-lite-v1","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"reasoning":0},"canonical_model_id":"amazon/nova-2-lite-v1"}}},"wafer.ai":{"id":"wafer.ai","env":["WAFER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://pass.wafer.ai/v1","name":"Wafer","doc":"https://docs.wafer.ai/wafer-pass","models":{"Kimi-K2.6":{"id":"Kimi-K2.6","name":"Kimi K2.6","description":"Kimi K2.6 sparse MoE model with a 262K context window. Available serverless and not included in standard Wafer Pass. Non-ZDR only: requests with `Wafer-ZDR: required` are rejected.","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":1.14,"output":4.8,"cache_read":0.19,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"MiniMax-M3":{"id":"MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"status":"deprecated","cost":{"input":0.33,"output":1.32,"cache_read":0.07,"cache_write":0,"tiers":[{"input":0.66,"output":2.64,"cache_read":0.13,"cache_write":0,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.66,"output":2.64,"cache_read":0.13,"cache_write":0}},"canonical_model_id":"minimax/MiniMax-M3"},"GLM-5.1":{"id":"GLM-5.1","name":"GLM-5.1","description":"General Language Model 5.1 — high-quality bilingual (EN/ZH) generation with strong coding and reasoning capabilities.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"status":"deprecated","cost":{"input":1,"output":3.2,"cache_read":0.1,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"glm5.2-fast":{"id":"glm5.2-fast","name":"GLM5.2-Fast","description":"The same model served for high TPS.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":3,"output":10.25,"cache_read":0.5,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"GLM-5.2":{"id":"GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"fireworks-ai":{"id":"fireworks-ai","env":["FIREWORKS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.fireworks.ai/inference/v1/","name":"Fireworks AI","doc":"https://fireworks.ai/docs/","models":{"accounts/fireworks/models/gpt-oss-120b":{"id":"accounts/fireworks/models/gpt-oss-120b","name":"GPT OSS 120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"status":"deprecated","experimental":{"modes":{"priority":{"cost":{"input":0.18,"output":0.72,"cache_read":0.018},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"openai/gpt-oss-120b"},"accounts/fireworks/models/glm-5p3":{"id":"accounts/fireworks/models/glm-5p3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":262144},"experimental":{"modes":{"priority":{"cost":{"input":1.75,"output":5.5,"cache_read":0.325},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/models/nemotron-3-ultra-nvfp4":{"id":"accounts/fireworks/models/nemotron-3-ultra-nvfp4","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"status":"deprecated","cost":{"input":0.6,"output":2.4,"cache_read":0.12},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"accounts/fireworks/models/minimax-m3":{"id":"accounts/fireworks/models/minimax-m3","name":"MiniMax-M3","description":"Fireworks text-only MiniMax coding model for long-context reasoning and agent tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"experimental":{"modes":{"priority":{"cost":{"input":0.45,"output":1.8,"cache_read":0.09},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"accounts/fireworks/models/deepseek-v4p1-flash":{"id":"accounts/fireworks/models/deepseek-v4p1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"experimental":{"modes":{"priority":{"cost":{"input":0.375,"output":1.5,"cache_read":0.0075},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"accounts/fireworks/models/kimi-k3":{"id":"accounts/fireworks/models/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3.75,"output":18.75,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/models/deepseek-v4p1-flash-1017":{"id":"accounts/fireworks/models/deepseek-v4p1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"experimental":{"modes":{"priority":{"cost":{"input":0.375,"output":1.5,"cache_read":0.0075},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"accounts/fireworks/models/ember-1":{"id":"accounts/fireworks/models/ember-1","name":"Ember-1","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"accounts/fireworks/models/qwen3p8-max":{"id":"accounts/fireworks/models/qwen3p8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3,"output":9,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"accounts/fireworks/models/qwen3p8-2p4t-a95b":{"id":"accounts/fireworks/models/qwen3p8-2p4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"accounts/fireworks/models/nemotron-lightning-3p5-30b-a3b":{"id":"accounts/fireworks/models/nemotron-lightning-3p5-30b-a3b","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","experimental":{"modes":{"priority":{"cost":{"input":0.0625,"output":0.25,"cache_read":0.0125},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"accounts/fireworks/models/minimax-m3p1-flash":{"id":"accounts/fireworks/models/minimax-m3p1-flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"accounts/fireworks/models/inkling":{"id":"accounts/fireworks/models/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"status":"deprecated","cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"accounts/fireworks/models/glm-5p3-flash":{"id":"accounts/fireworks/models/glm-5p3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":0.1875,"output":0.625,"cache_read":0.0375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"accounts/fireworks/models/muse-spark-open":{"id":"accounts/fireworks/models/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"accounts/fireworks/routers/kimi-latest":{"id":"accounts/fireworks/routers/kimi-latest","name":"Kimi Latest","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3.75,"output":18.75,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/qwen-max-latest":{"id":"accounts/fireworks/routers/qwen-max-latest","name":"Qwen Max Latest (Qwen3.8 Max)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":3,"output":9,"cache_read":0.375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"accounts/fireworks/routers/deepseek-flash-latest":{"id":"accounts/fireworks/routers/deepseek-flash-latest","name":"DeepSeek Flash Latest","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"experimental":{"modes":{"priority":{"cost":{"input":0.375,"output":1.5,"cache_read":0.0075},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"accounts/fireworks/routers/glm-flash-latest":{"id":"accounts/fireworks/routers/glm-flash-latest","name":"GLM Flash Latest (GLM 5.3 Flash)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":131072},"experimental":{"modes":{"priority":{"cost":{"input":0.1875,"output":0.625,"cache_read":0.0375},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"accounts/fireworks/routers/glm-5p3-fast":{"id":"accounts/fireworks/routers/glm-5p3-fast","name":"GLM 5.3 Fast","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048572,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.39},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/routers/glm-5p3-spot":{"id":"accounts/fireworks/routers/glm-5p3-spot","name":"GLM 5.3 Spot","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":262144},"cost":{"input":0.98,"output":3.08,"cache_read":0.182},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/routers/deepseek-v4p1-flash-1017-spot":{"id":"accounts/fireworks/routers/deepseek-v4p1-flash-1017-spot","name":"DeepSeek V4.1 Flash 1017 Spot","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.21,"output":0.84,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"accounts/fireworks/routers/minimax-latest":{"id":"accounts/fireworks/routers/minimax-latest","name":"MiniMax Latest","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"output":512000},"experimental":{"modes":{"priority":{"cost":{"input":0.45,"output":1.8,"cache_read":0.09},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"accounts/fireworks/routers/kimi-k3-spot":{"id":"accounts/fireworks/routers/kimi-k3-spot","name":"Kimi K3 Spot","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":2.1,"output":10.5,"cache_read":0.21},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/deepseek-v4p1-flash-spot":{"id":"accounts/fireworks/routers/deepseek-v4p1-flash-spot","name":"DeepSeek V4.1 Flash Spot","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.21,"output":0.84,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"accounts/fireworks/routers/kimi-k3-fast":{"id":"accounts/fireworks/routers/kimi-k3-fast","name":"Kimi K3 Fast","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/kimi-fast-latest":{"id":"accounts/fireworks/routers/kimi-fast-latest","name":"Kimi Fast Latest","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.5,"output":22.5,"cache_read":0.45},"canonical_model_id":"moonshotai/kimi-k3"},"accounts/fireworks/routers/glm-latest":{"id":"accounts/fireworks/routers/glm-latest","name":"GLM Latest","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048573,"output":262144},"experimental":{"modes":{"priority":{"cost":{"input":1.75,"output":5.5,"cache_read":0.325},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"accounts/fireworks/routers/glm-fast-latest":{"id":"accounts/fireworks/routers/glm-fast-latest","name":"GLM 5.3 Fast (Latest)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048572,"output":262144},"cost":{"input":2.1,"output":6.6,"cache_read":0.39},"canonical_model_id":"zhipuai/glm-5.3"}}},"helicone":{"id":"helicone","env":["HELICONE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://ai-gateway.helicone.ai/v1","name":"Helicone","doc":"https://helicone.ai/models","models":{"claude-3-haiku-20240307":{"id":"claude-3-haiku-20240307","name":"Anthropic: Claude 3 Haiku","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2024-03-13","last_updated":"2024-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":4096},"cost":{"input":0.25,"output":1.25,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"anthropic/claude-3-haiku-20240307"},"kimi-k2-0905":{"id":"kimi-k2-0905","name":"Kimi K2 (09/05)","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.5,"output":2,"cache_read":0.4},"canonical_model_id":"moonshotai/kimi-k2-0905"},"qwen3-235b-a22b-thinking":{"id":"qwen3-235b-a22b-thinking","name":"Qwen3 235B A22B Thinking","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.3,"output":2.9},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen3-vl-235b-a22b-instruct":{"id":"qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.3,"output":1.5},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"o1":{"id":"o1","name":"OpenAI: o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"glm-4.6":{"id":"glm-4.6","name":"Zai GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.45,"output":1.5},"canonical_model_id":"zhipuai/glm-4.6"},"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"OpenAI GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"hermes-2-pro-llama-3-8b":{"id":"hermes-2-pro-llama-3-8b","name":"Hermes 2 Pro Llama 3 8B","description":"Llama 3 fine-tune optimized for function calling, structured JSON responses, and dialogue","family":"hermes","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-05","release_date":"2024-04-30","last_updated":"2024-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.14,"output":0.14},"canonical_model_id":"nousresearch/hermes-2-pro-llama-3-8b"},"claude-sonnet-4":{"id":"claude-sonnet-4","name":"Anthropic: Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-0"},"claude-3.5-haiku":{"id":"claude-3.5-haiku","name":"Anthropic: Claude 3.5 Haiku","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":0.8,"output":4,"cache_read":0.08,"cache_write":1},"canonical_model_id":"anthropic/claude-3-5-haiku-20241022"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"OpenAI GPT-4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"OpenAI GPT-OSS 120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.04,"output":0.16},"canonical_model_id":"openai/gpt-oss-120b"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Meta Llama 3.3 70B Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16400},"cost":{"input":0.13,"output":0.39},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gpt-5.1":{"id":"gpt-5.1","name":"OpenAI GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"sonar":{"id":"sonar","name":"Perplexity Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"gemma-3-12b-it":{"id":"gemma-3-12b-it","name":"Google Gemma 3 12B","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-12","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.05,"output":0.1},"canonical_model_id":"google/gemma-3-12b-it"},"grok-3-mini":{"id":"grok-3-mini","name":"xAI Grok 3 Mini","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":0.3,"output":0.5,"cache_read":0.075}},"qwen3-coder":{"id":"qwen3-coder","name":"Qwen3 Coder 480B A35B Instruct Turbo","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.22,"output":0.95},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"kimi-k2-0711":{"id":"kimi-k2-0711","name":"Kimi K2 (07/11)","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.57,"output":2.3},"canonical_model_id":"moonshotai/kimi-k2"},"qwen2.5-coder-7b-fast":{"id":"qwen2.5-coder-7b-fast","name":"Qwen2.5 Coder 7B fast","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-09","release_date":"2024-09-15","last_updated":"2024-09-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192},"cost":{"input":0.03,"output":0.09}},"qwen3-coder-30b-a3b-instruct":{"id":"qwen3-coder-30b-a3b-instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"llama-4-maverick":{"id":"llama-4-maverick","name":"Meta Llama 4 Maverick 17B 128E","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"deepseek-v3.1-terminus":{"id":"deepseek-v3.1-terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.27,"output":1,"cache_read":0.216}},"gpt-4.1":{"id":"gpt-4.1","name":"OpenAI GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"OpenAI: GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"ernie-4.5-21b-a3b-thinking":{"id":"ernie-4.5-21b-a3b-thinking","name":"Baidu Ernie 4.5 21B A3B Thinking","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"ernie","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-03","release_date":"2025-03-16","last_updated":"2025-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8000},"cost":{"input":0.07,"output":0.28}},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"OpenAI GPT-OSS 20b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.05,"output":0.2},"canonical_model_id":"openai/gpt-oss-20b"},"grok-4-1-fast-reasoning":{"id":"grok-4-1-fast-reasoning","name":"xAI Grok 4.1 Fast Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast-reasoning"},"deepseek-reasoner":{"id":"deepseek-reasoner","name":"DeepSeek Reasoner","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":0.56,"output":1.68,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-reasoner"},"claude-3.7-sonnet":{"id":"claude-3.7-sonnet","name":"Anthropic: Claude 3.7 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"Anthropic: Claude Opus 4.1 (20250805)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-08","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gpt-5-mini":{"id":"gpt-5-mini","name":"OpenAI GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"mistral-small":{"id":"mistral-small","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.075,"output":0.2},"canonical_model_id":"mistral/mistral-small-2506"},"claude-4.5-haiku":{"id":"claude-4.5-haiku","name":"Anthropic: Claude 4.5 Haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"xAI Grok 4 Fast Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"chatgpt-4o-latest":{"id":"chatgpt-4o-latest","name":"OpenAI ChatGPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2024-08-14","last_updated":"2024-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":5,"output":20,"cache_read":2.5}},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Google Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.3,"output":2.5,"cache_read":0.075,"cache_write":0.3},"canonical_model_id":"google/gemini-2.5-flash"},"claude-4.5-sonnet":{"id":"claude-4.5-sonnet","name":"Anthropic: Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"llama-3.1-8b-instruct":{"id":"llama-3.1-8b-instruct","name":"Meta Llama 3.1 8B Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":16384},"cost":{"input":0.02,"output":0.05},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"sonar-reasoning-pro":{"id":"sonar-reasoning-pro","name":"Perplexity Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-reasoning-pro"},"claude-4.5-opus":{"id":"claude-4.5-opus","name":"Anthropic: Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"o1-mini":{"id":"o1-mini","name":"OpenAI: o1-mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":65536},"cost":{"input":1.1,"output":4.4,"cache_read":0.55}},"gpt-5-nano":{"id":"gpt-5-nano","name":"OpenAI GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"claude-opus-4-1":{"id":"claude-opus-4-1","name":"Anthropic: Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-08","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-1"},"llama-3.3-70b-versatile":{"id":"llama-3.3-70b-versatile","name":"Meta Llama 3.3 70B Versatile","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32678},"cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"qwen3-next-80b-a3b-instruct":{"id":"qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":16384},"cost":{"input":0.14,"output":1.4},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"xAI Grok Code Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":10000},"cost":{"input":0.2,"output":1.5,"cache_read":0.02},"canonical_model_id":"xai/grok-code-fast-1"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Google Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.3125,"cache_write":1.25},"canonical_model_id":"google/gemini-2.5-pro"},"o3-mini":{"id":"o3-mini","name":"OpenAI o3 Mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2023-10","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"grok-4":{"id":"grok-4","name":"xAI Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-09","last_updated":"2024-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":3,"output":15,"cache_read":0.75}},"sonar-reasoning":{"id":"sonar-reasoning","name":"Perplexity Sonar Reasoning","description":"Web-grounded reasoning model for multi-step research and cited answers","family":"sonar-reasoning","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-01-27","last_updated":"2025-01-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":1,"output":5}},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Anthropic: Claude Sonnet 4.5 (20250929)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":63999}],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"llama-prompt-guard-2-86m":{"id":"llama-prompt-guard-2-86m","name":"Meta Llama Prompt Guard 2 86M","description":"86M-parameter (mDeBERTa-base) multilingual classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":2},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"meta/llama-prompt-guard-2-86m"},"gpt-5-pro":{"id":"gpt-5-pro","name":"OpenAI: GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":32768},"cost":{"input":15,"output":120},"canonical_model_id":"openai/gpt-5-pro"},"llama-4-scout":{"id":"llama-4-scout","name":"Meta Llama 4 Scout 17B 16E","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.08,"output":0.3},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"sonar-deep-research":{"id":"sonar-deep-research","name":"Perplexity Sonar Deep Research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":127000,"output":4096},"cost":{"input":2,"output":8},"canonical_model_id":"perplexity/sonar-deep-research"},"o4-mini":{"id":"o4-mini","name":"OpenAI o4 Mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"qwen3-32b":{"id":"qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":40960},"cost":{"input":0.29,"output":0.59},"canonical_model_id":"alibaba/qwen3-32b"},"claude-opus-4":{"id":"claude-opus-4","name":"Anthropic: Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024,"max":31999}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-0"},"gemma2-9b-it":{"id":"gemma2-9b-it","name":"Google Gemma 2","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2024-06-25","last_updated":"2024-06-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"cost":{"input":0.01,"output":0.03}},"llama-3.1-8b-instant":{"id":"llama-3.1-8b-instant","name":"Meta Llama 3.1 8B Instant","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32678},"cost":{"input":0.05,"output":0.08},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"deepseek-r1-distill-llama-70b":{"id":"deepseek-r1-distill-llama-70b","name":"DeepSeek R1 Distill Llama 70B","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-01-20","last_updated":"2025-01-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":0.03,"output":0.13}},"gpt-5.1-chat-latest":{"id":"gpt-5.1-chat-latest","name":"OpenAI GPT-5.1 Chat","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"grok-3":{"id":"grok-3","name":"xAI Grok 3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.75}},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"OpenAI: GPT-5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"gpt-4o":{"id":"gpt-4o","name":"OpenAI GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-05","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"xAI Grok 4.1 Fast Non-Reasoning","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4.1-fast"},"llama-prompt-guard-2-22m":{"id":"llama-prompt-guard-2-22m","name":"Meta Llama Prompt Guard 2 22M","description":"22M-parameter (DeBERTa-xsmall) English classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":2},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"meta/llama-prompt-guard-2-22m"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Anthropic: Claude 4.5 Haiku (20251001)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gpt-4.1-mini-2025-04-14":{"id":"gpt-4.1-mini-2025-04-14","name":"OpenAI GPT-4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"o3":{"id":"o3","name":"OpenAI o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"deepseek-v3":{"id":"deepseek-v3","name":"DeepSeek V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.56,"output":1.68,"cache_read":0.07}},"grok-4-fast-reasoning":{"id":"grok-4-fast-reasoning","name":"xAI: Grok 4 Fast Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000},"cost":{"input":0.2,"output":0.5,"cache_read":0.05},"canonical_model_id":"xai/grok-4-fast-reasoning"},"sonar-pro":{"id":"sonar-pro","name":"Perplexity Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":4096},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"kimi-k2-thinking":{"id":"kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":262144},"cost":{"input":0.48,"output":2},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"mistral-large-2411":{"id":"mistral-large-2411","name":"Mistral-Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-11-18","last_updated":"2024-11-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":6},"canonical_model_id":"mistral/mistral-large-2411"},"llama-3.1-8b-instruct-turbo":{"id":"llama-3.1-8b-instruct-turbo","name":"Meta Llama 3.1 8B Instruct Turbo","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.02,"output":0.03},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"deepseek-v3.2":{"id":"deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":65536},"cost":{"input":0.27,"output":0.41},"canonical_model_id":"deepseek/deepseek-v3.2"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"OpenAI GPT-4o-mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"claude-3.5-sonnet-v2":{"id":"claude-3.5-sonnet-v2","name":"Anthropic: Claude 3.5 Sonnet v2","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-3-5-sonnet-20241022"},"llama-guard-4":{"id":"llama-guard-4","name":"Meta Llama Guard 4 12B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-01-01","last_updated":"2025-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":1024},"cost":{"input":0.21,"output":0.21}},"gpt-5":{"id":"gpt-5","name":"OpenAI GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"Google Gemini 3 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-preview"},"o3-pro":{"id":"o3-pro","name":"OpenAI o3 Pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2024-06","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"qwen3-30b-a3b":{"id":"qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"output":41000},"cost":{"input":0.08,"output":0.29},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"mistral-nemo":{"id":"mistral-nemo","name":"Mistral Nemo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16400},"cost":{"input":20,"output":40},"canonical_model_id":"mistral/mistral-nemo"},"gpt-5-codex":{"id":"gpt-5-codex","name":"OpenAI: GPT-5 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Google Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65535},"cost":{"input":0.1,"output":0.4,"cache_read":0.025,"cache_write":0.1},"canonical_model_id":"google/gemini-2.5-flash-lite"},"deepseek-tng-r1t2-chimera":{"id":"deepseek-tng-r1t2-chimera","name":"DeepSeek TNG R1T2 Chimera","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek-thinking","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2025-07-02","last_updated":"2025-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":130000,"output":163840},"cost":{"input":0.3,"output":1.2}},"gpt-5-chat-latest":{"id":"gpt-5-chat-latest","name":"OpenAI GPT-5 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":false,"knowledge":"2024-09","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":272000,"output":16384},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-chat-latest"}}},"gitlab":{"id":"gitlab","env":["GITLAB_TOKEN"],"npm":"gitlab-ai-provider","name":"GitLab Duo","doc":"https://docs.gitlab.com/user/duo_agent_platform/","models":{"duo-chat-gpt-5-2":{"id":"duo-chat-gpt-5-2","name":"Agentic Chat (GPT-5.2)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.2"},"duo-chat-gpt-5-2-codex":{"id":"duo-chat-gpt-5-2-codex","name":"Agentic Chat (GPT-5.2 Codex)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.2-codex"},"duo-chat-gpt-5-4":{"id":"duo-chat-gpt-5-4","name":"Agentic Chat (GPT-5.4)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4"},"duo-chat-opus-4-8":{"id":"duo-chat-opus-4-8","name":"Agentic Chat (Claude Opus 4.8)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-8"},"duo-chat-sonnet-4-5":{"id":"duo-chat-sonnet-4-5","name":"Agentic Chat (Claude Sonnet 4.5)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"duo-chat-opus-5":{"id":"duo-chat-opus-5","name":"Agentic Chat (Claude Opus 5)","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-5"},"duo-chat-gpt-5-6-sol":{"id":"duo-chat-gpt-5-6-sol","name":"Agentic Chat (GPT-5.6 Sol)","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-5.6-sol"},"duo-chat-sonnet-4-6":{"id":"duo-chat-sonnet-4-6","name":"Agentic Chat (Claude Sonnet 4.6)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"duo-chat-gpt-5-3-codex":{"id":"duo-chat-gpt-5-3-codex","name":"Agentic Chat (GPT-5.3 Codex)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.3-codex"},"duo-chat-opus-4-6":{"id":"duo-chat-opus-4-6","name":"Agentic Chat (Claude Opus 4.6)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-6"},"duo-chat-gpt-6-astra":{"id":"duo-chat-gpt-6-astra","name":"Agentic Chat (GPT-6 Astra)","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6-astra"},"duo-chat-sonnet-5":{"id":"duo-chat-sonnet-5","name":"Agentic Chat (Claude Sonnet 5)","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-5"},"duo-chat-gpt-5-4-mini":{"id":"duo-chat-gpt-5-4-mini","name":"Agentic Chat (GPT-5.4 Mini)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4-mini"},"duo-chat-gpt-5-codex":{"id":"duo-chat-gpt-5-codex","name":"Agentic Chat (GPT-5 Codex)","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5-codex"},"duo-chat-haiku-4-5":{"id":"duo-chat-haiku-4-5","name":"Agentic Chat (Claude Haiku 4.5)","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"duo-chat-gpt-5-4-nano":{"id":"duo-chat-gpt-5-4-nano","name":"Agentic Chat (GPT-5.4 Nano)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4-nano"},"duo-chat-gpt-5-6-luna":{"id":"duo-chat-gpt-5-6-luna","name":"Agentic Chat (GPT-5.6 Luna)","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-5.6-luna"},"duo-chat-sonnet-5-5":{"id":"duo-chat-sonnet-5-5","name":"Agentic Chat (Claude Sonnet 5.5)","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"duo-chat-opus-4-5":{"id":"duo-chat-opus-4-5","name":"Agentic Chat (Claude Opus 4.5)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"duo-chat-gpt-5-5":{"id":"duo-chat-gpt-5-5","name":"Agentic Chat (GPT-5.5)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.5"},"duo-chat-opus-5-5":{"id":"duo-chat-opus-5-5","name":"Agentic Chat (Claude Opus 5.5)","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-5-5"},"duo-chat-gpt-5-1":{"id":"duo-chat-gpt-5-1","name":"Agentic Chat (GPT-5.1)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.1"},"duo-chat-opus-4-7":{"id":"duo-chat-opus-4-7","name":"Agentic Chat (Claude Opus 4.7)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-opus-4-7"},"duo-chat-fable-5":{"id":"duo-chat-fable-5","name":"Agentic Chat (Claude Fable 5)","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-fable-5"},"duo-chat-gpt-6-luna":{"id":"duo-chat-gpt-6-luna","name":"Agentic Chat (GPT-6 Luna)","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6-luna"},"duo-chat-gpt-6-sol":{"id":"duo-chat-gpt-6-sol","name":"Agentic Chat (GPT-6 Sol)","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6-sol"},"duo-chat-gpt-5-mini":{"id":"duo-chat-gpt-5-mini","name":"Agentic Chat (GPT-5 Mini)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5-mini"},"duo-chat-fable-5-1":{"id":"duo-chat-fable-5-1","name":"Agentic Chat (Claude Fable 5.1)","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"anthropic/claude-fable-5-1"},"duo-chat-gpt-6-1-sol":{"id":"duo-chat-gpt-6-1-sol","name":"Agentic Chat (GPT-6.1 Sol)","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-6.1-sol"},"duo-chat-gpt-5-6-terra":{"id":"duo-chat-gpt-5-6-terra","name":"Agentic Chat (GPT-5.6 Terra)","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"openai/gpt-5.6-terra"}}},"longcat":{"id":"longcat","env":["LONGCAT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.longcat.chat/openai","name":"LongCat","doc":"https://longcat.chat/platform/docs/","models":{"LongCat-2.0":{"id":"LongCat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.0"},"LongCat-2.5-Preview":{"id":"LongCat-2.5-Preview","name":"LongCat-2.5-Preview","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"meituan/longcat-2.5-preview"}}},"sap-ai-core":{"id":"sap-ai-core","env":["AICORE_SERVICE_KEY"],"npm":"@jerome-benoit/sap-ai-provider-v2","name":"SAP AI Core","doc":"https://help.sap.com/docs/sap-ai-core","models":{"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"gpt-4.1-mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"gpt-4.1-nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":0.1,"output":0.4},"canonical_model_id":"openai/gpt-4.1-nano"},"nvidia--llama-3.2-nv-embedqa-1b":{"id":"nvidia--llama-3.2-nv-embedqa-1b","name":"nvidia--llama-3.2-nv-embedqa-1b","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":4096},"status":"deprecated"},"sonar":{"id":"sonar","name":"sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"cost":{"input":1,"output":1},"canonical_model_id":"perplexity/sonar"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"gemini-3.5-flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"gpt-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-6-astra"},"sap-abap-1":{"id":"sap-abap-1","name":"sap-abap-1","description":"SAP-hosted model for ABAP code generation and enterprise development tasks","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-11-26","last_updated":"2025-11-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":4096},"status":"deprecated"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3.8-flash"},"anthropic--claude-4.6-sonnet":{"id":"anthropic--claude-4.6-sonnet","name":"anthropic--claude-4.6-sonnet","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gpt-4.1":{"id":"gpt-4.1","name":"gpt-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"status":"deprecated","cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"anthropic--claude-4.5-opus":{"id":"anthropic--claude-4.5-opus","name":"anthropic--claude-4.5-opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"gpt-5-mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"anthropic--claude-3-haiku":{"id":"anthropic--claude-3-haiku","name":"anthropic--claude-3-haiku","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-08-31","release_date":"2024-03-13","last_updated":"2024-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":4096},"cost":{"input":0.25,"output":1.25,"cache_read":0.03,"cache_write":0.3},"canonical_model_id":"anthropic/claude-3-haiku-20240307"},"gpt-5.5":{"id":"gpt-5.5","name":"gpt-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"gemini-2.5-flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"cohere--command-a-reasoning":{"id":"cohere--command-a-reasoning","name":"cohere--command-a-reasoning","description":"Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]},{"type":"budget_tokens","min":1}],"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-reasoning-08-2025"},"amazon--nova-lite":{"id":"amazon--nova-lite","name":"amazon--nova-lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"amazon/nova-2-lite"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-6-luna"},"gpt-5-nano":{"id":"gpt-5-nano","name":"gpt-5-nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"gemini-2.5-pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"google/gemini-2.5-pro"},"anthropic--claude-4.5-sonnet":{"id":"anthropic--claude-4.5-sonnet","name":"anthropic--claude-4.5-sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic--claude-4.6-opus":{"id":"anthropic--claude-4.6-opus","name":"anthropic--claude-4.6-opus","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"amazon--titan-embed-text":{"id":"amazon--titan-embed-text","name":"amazon--titan-embed-text","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-04-30","last_updated":"2024-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":1536},"cost":{"input":0.02,"output":0}},"gpt-5.2":{"id":"gpt-5.2","name":"gpt-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"gpt-5.6-terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"openai/gpt-5.6-terra"},"anthropic--claude-4.5-haiku":{"id":"anthropic--claude-4.5-haiku","name":"anthropic--claude-4.5-haiku","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic--claude-4-opus":{"id":"anthropic--claude-4-opus","name":"anthropic--claude-4-opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75,"cache_read":1.5,"cache_write":18.75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"text-embedding-3-small":{"id":"text-embedding-3-small","type":"embedding","name":"text-embedding-3-small","description":"Small, efficient OpenAI text embedding model","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":1536},"cost":{"input":0.02,"output":0},"canonical_model_id":"openai/text-embedding-3-small"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"canonical_model_id":"openai/gpt-6.1-sol"},"mistralai--mistral-medium":{"id":"mistralai--mistral-medium","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"canonical_model_id":"mistral/mistral-medium-2604"},"gemini-embedding":{"id":"gemini-embedding","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1},"canonical_model_id":"google/gemini-embedding-001"},"sonar-deep-research":{"id":"sonar-deep-research","name":"sonar-deep-research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768},"cost":{"input":2,"output":8,"reasoning":3},"canonical_model_id":"perplexity/sonar-deep-research"},"nvidia--llama-nemotron-embed-1b":{"id":"nvidia--llama-nemotron-embed-1b","type":"embedding","name":"Llama Nemotron Embed 1B v2","description":"Multilingual text embedding model for retrieval and question answering pipelines","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"canonical_model_id":"nvidia/llama-nemotron-embed-1b-v2"},"amazon--nova-pro":{"id":"amazon--nova-pro","name":"amazon--nova-pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":8192},"cost":{"input":0.8,"output":3.2},"canonical_model_id":"amazon/nova-pro"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3.6-flash"},"mistralai--mistral-small":{"id":"mistralai--mistral-small","name":"mistralai--mistral-small","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.15,"output":0.6},"canonical_model_id":"mistral/mistral-small-2603"},"anthropic--claude-4.7-opus":{"id":"anthropic--claude-4.7-opus","name":"anthropic--claude-4.7-opus","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"sonar-pro":{"id":"sonar-pro","name":"sonar-pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15},"canonical_model_id":"perplexity/sonar-pro"},"amazon--nova-micro":{"id":"amazon--nova-micro","name":"amazon--nova-micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192},"cost":{"input":0.035,"output":0.14},"canonical_model_id":"amazon/nova-micro"},"mistralai--mistral-medium-instruct":{"id":"mistralai--mistral-medium-instruct","name":"mistralai--mistral-medium-instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"status":"deprecated","cost":{"input":0.4,"output":2},"canonical_model_id":"mistral/mistral-medium-2505"},"text-embedding-3-large":{"id":"text-embedding-3-large","type":"embedding","name":"text-embedding-3-large","description":"Most capable OpenAI embedding model for English and non-English text","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-25","last_updated":"2024-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8191,"output":3072},"cost":{"input":0.13,"output":0},"canonical_model_id":"openai/text-embedding-3-large"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"gpt-5.6-luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04}},"canonical_model_id":"openai/gpt-5.6-luna"},"gemini-embedding-2":{"id":"gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072},"canonical_model_id":"google/gemini-embedding-2"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gpt-5":{"id":"gpt-5","name":"gpt-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"anthropic--claude-4-sonnet":{"id":"anthropic--claude-4-sonnet","name":"anthropic--claude-4-sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"gemini-3.1-flash-lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"gpt-5.6-sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"tiers":[{"input":8,"output":30,"cache_read":0.8,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8}},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"gemini-2.5-flash-lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"anthropic--claude-4.8-opus":{"id":"anthropic--claude-4.8-opus","name":"anthropic--claude-4.8-opus","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"}}},"regolo-ai":{"id":"regolo-ai","env":["REGOLO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.regolo.ai/v1","name":"Regolo AI","doc":"https://docs.regolo.ai/","models":{"qwen3.5-9b":{"id":"qwen3.5-9b","name":"Qwen3.5-9B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.078505,"output":0.392525},"canonical_model_id":"alibaba/qwen3.5-9b"},"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT-OSS-120B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":1.1215,"output":4.7103},"canonical_model_id":"openai/gpt-oss-120b"},"llama-3.3-70b-instruct":{"id":"llama-3.3-70b-instruct","name":"Llama 3.3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":0.6,"output":2.7},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"faster-whisper-large-v3":{"id":"faster-whisper-large-v3","type":"transcription","name":"Faster Whisper Large v3","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"openai/whisper-large-v3"},"gpt-oss-20b":{"id":"gpt-oss-20b","name":"GPT-OSS-20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.11215,"output":0.47103},"canonical_model_id":"openai/gpt-oss-20b"},"mistral-small-4-119b":{"id":"mistral-small-4-119b","name":"Mistral Small 4 119B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.56075,"output":2.35515},"canonical_model_id":"mistral/mistral-small-2603"},"qwen3-coder-next":{"id":"qwen3-coder-next","name":"Qwen3-Coder-Next","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.3,"output":1.2},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen3.5-122b":{"id":"qwen3.5-122b","name":"Qwen3.5-122B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":1.1215,"output":4.7103},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"glm5.2":{"id":"glm5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":96000,"output":96000},"cost":{"input":2.31,"output":6},"canonical_model_id":"zhipuai/glm-5.2"},"qwen3-embedding-8b":{"id":"qwen3-embedding-8b","type":"embedding","name":"Qwen3-Embedding-8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.1,"output":0.1},"canonical_model_id":"alibaba/qwen3-embedding-8b"},"gemma4-31b":{"id":"gemma4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":100000,"output":100000},"cost":{"input":0.46,"output":2.42},"canonical_model_id":"google/gemma-4-31b-it"},"qwen-image":{"id":"qwen-image","type":"image","name":"Qwen-Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-01","last_updated":"2026-03-01","modalities":{"input":["text"],"output":["image"]},"open_weights":false,"limit":{"context":8192,"output":4096},"cost":{"input":0.5,"output":2}},"brick-v1-beta":{"id":"brick-v1-beta","name":"Brick v1 Beta","description":"Semantic router by Regolo.ai that directs each request to the most suitable model, optimizing costs and performance","family":"model-router","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-06","last_updated":"2026-02-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":100000,"output":15000},"status":"beta","cost":{"input":0,"output":0}},"deepseek-ocr-2":{"id":"deepseek-ocr-2","name":"DeepSeek OCR 2","description":"High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":4000,"output":4000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-ocr-2"},"qwen3-reranker-4b":{"id":"qwen3-reranker-4b","type":"reranking","name":"Qwen3-Reranker-4B","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-01","last_updated":"2026-02-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"cost":{"input":0.12,"output":0.12}},"apertus-70b":{"id":"apertus-70b","name":"Apertus 70B","description":"Fully open 70B multilingual LLM supporting 1800+ languages with 65K context. Trained on 15T tokens of compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":30000,"output":30000},"cost":{"input":0.46,"output":2.42},"canonical_model_id":"swiss-ai/apertus-70b"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":120000,"output":120000},"cost":{"input":0.58,"output":2.42},"canonical_model_id":"alibaba/qwen3.8-27b"},"brick-complexity-pro":{"id":"brick-complexity-pro","name":"Brick Complexity Pro","description":"Complexity classifier that powers the Brick semantic router by extracting query difficulty","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-06","last_updated":"2026-02-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":100000,"output":15000},"cost":{"input":0.12,"output":0.46}}}},"aihubmix":{"id":"aihubmix","env":["AIHUBMIX_API_KEY"],"npm":"@aihubmix/ai-sdk-provider","name":"AIHubMix","doc":"https://docs.aihubmix.com","models":{"glm-5v-turbo":{"id":"glm-5v-turbo","name":"GLM 5 Vision Turbo","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":0.7042,"output":3.09848,"cache_read":0.169008,"tiers":[{"input":0.986,"output":3.662004,"cache_read":0.253402,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5v-turbo"},"gpt-image-1.5":{"id":"gpt-image-1.5","type":"image","name":"GPT-Image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":32},"canonical_model_id":"openai/gpt-image-1.5"},"mimo-v2.6-pro":{"id":"mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.48,"output":0.96,"cache_read":0.00384},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"glm-4.5v":{"id":"glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"cost":{"input":0.274,"output":1.096,"tiers":[{"input":0.548,"output":1.644,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.5v"},"glm-4.6":{"id":"glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.273974,"output":1.095896,"cache_read":0.0547948,"tiers":[{"input":0.547946,"output":2.191784,"cache_read":0.1095892,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6"},"xiaomi-mimo-v2.5":{"id":"xiaomi-mimo-v2.5","name":"Xiaomi MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.44,"output":2.2,"cache_read":0.088,"tiers":[{"input":0.88,"output":4.4,"cache_read":0.176,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.88,"output":4.4,"cache_read":0.176}},"canonical_model_id":"xiaomi/mimo-v2.5"},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":false,"interleaved":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":3.9995,"cache_read":0.160835},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"claude-opus-4-7-think":{"id":"claude-opus-4-7-think","name":"Claude Opus 4.7 Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"coding-glm-5.3-free":{"id":"coding-glm-5.3-free","name":"GLM-5.3 (Coding, free)","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3"},"claude-sonnet-4-5-think":{"id":"claude-sonnet-4-5-think","name":"Claude Sonnet 4.5 Thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1268,"output":3.9438,"cache_read":0.2817},"canonical_model_id":"zhipuai/glm-5.2"},"claude-haiku-5-5":{"id":"claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"nemotron-3-ultra-550b-a55b-free":{"id":"nemotron-3-ultra-550b-a55b-free","name":"Nemotron 3 Ultra 550B A55B (free)","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"qwen3.7-plus":{"id":"qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":64000},"cost":{"input":0.282,"output":1.128,"cache_read":0.0564,"cache_write":0.3525,"tiers":[{"input":0.845,"output":3.38,"cache_read":0.169,"cache_write":1.05625,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.845,"output":3.38,"cache_read":0.169,"cache_write":1.05625}},"canonical_model_id":"alibaba/qwen3.7-plus"},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05,"input_audio":1},"canonical_model_id":"google/gemini-3-flash-preview"},"claude-sonnet-4-5":{"id":"claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3.3,"output":16.5,"cache_read":0.33,"cache_write":4.125,"tiers":[{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6.6,"output":24.75,"cache_read":0.66,"cache_write":8.25}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"gpt-image-2.5-flare":{"id":"gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-flare"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":5,"output":30,"cache_read":0.5},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"coding-glm-5.1":{"id":"coding-glm-5.1","name":"Coding GLM 5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.06,"output":0.22,"cache_read":0.013},"canonical_model_id":"zhipuai/glm-5.1"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"deep-deepseek-v4-flash":{"id":"deep-deepseek-v4-flash","name":"DeepSeek V4 Flash (DeepSeek)","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.154,"output":0.308,"cache_read":0.0308},"canonical_model_id":"deepseek/deepseek-v4-flash"},"coding-xiaomi-mimo-v2.5-pro":{"id":"coding-xiaomi-mimo-v2.5-pro","name":"Coding Xiaomi MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.2,"output":0.4,"cache_read":0.0016},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.155,"output":0.62,"cache_read":0.0031},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"qwen3.8-max-preview":{"id":"qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.338,"output":1.014,"cache_read":0.0676,"cache_write":0.4225},"canonical_model_id":"alibaba/qwen3.8-max-preview"},"step-3.7-flash":{"id":"step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"cost":{"input":0.22,"output":1.32,"cache_read":0.044},"canonical_model_id":"stepfun/step-3.7-flash"},"glm-4.7":{"id":"glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0.273974,"output":1.095896,"cache_read":0.0547948,"tiers":[{"input":0.547946,"output":2.191784,"cache_read":0.1095892,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.7"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview"},"alicloud-glm-5.1":{"id":"alicloud-glm-5.1","name":"GLM-5.1 (Alibaba Cloud)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":0.84,"output":3.38,"cache_read":0.169,"cache_write":1.05625},"canonical_model_id":"zhipuai/glm-5.1"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"gemini-3.1-pro-preview-customtools":{"id":"gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"tiers":[{"input":4,"output":18,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4}},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"doubao-seed-2-0-code-preview":{"id":"doubao-seed-2-0-code-preview","name":"Doubao Seed 2.0 Code Preview","description":"Coding model for repository understanding, refactors, and agentic engineering tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4822,"output":2.411,"cache_read":0.09644,"tiers":[{"input":0.72328,"output":3.6164,"cache_read":0.144656,"tier":{"type":"context","size":32000}},{"input":1.4466,"output":7.233,"cache_read":0.28932,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.7-flash"},"nemotron-3-super-120b-a12b-free":{"id":"nemotron-3-super-120b-a12b-free","name":"Nemotron 3 Super 120B A12B (free)","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"zai-glm-5.1":{"id":"zai-glm-5.1","name":"GLM-5.1 (Z.ai)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":128000},"status":"deprecated","cost":{"input":0.845,"output":3.38,"cache_read":0.183112},"canonical_model_id":"zhipuai/glm-5.1"},"coding-minimax-m2.7":{"id":"coding-minimax-m2.7","name":"Coding MiniMax M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128100},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"hy3-preview":{"id":"hy3-preview","name":"Hy3 Preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":128000},"status":"deprecated","cost":{"input":0.17,"output":0.566661,"cache_read":0.051},"canonical_model_id":"tencent/hy3-preview"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-opus-4-8-think":{"id":"claude-opus-4-8-think","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"doubao-seed-2-0-lite-260428":{"id":"doubao-seed-2-0-lite-260428","name":"Doubao Seed 2.0 Lite 260428","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.09041,"output":0.54246,"cache_read":0.018082,"input_audio":1.35615,"tiers":[{"input":0.135616,"output":0.813696,"cache_read":0.0271232,"tier":{"type":"context","size":32000}},{"input":0.271232,"output":1.627392,"cache_read":0.0542464,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.845,"output":2.535,"cache_read":0.04225},"canonical_model_id":"tencent/hy4-preview"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"gemini-3.1-flash-lite-image":{"id":"gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","high"]}],"tool_call":true,"interleaved":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096},"cost":{"input":0.25,"output":1.5,"cache_read":0.025},"canonical_model_id":"google/gemini-3.1-flash-lite-image"},"qwen3-vl-plus":{"id":"qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768},"cost":{"input":0.137,"output":1.37,"cache_read":0,"tiers":[{"input":0.2054,"output":2.054,"tier":{"type":"context","size":32000}},{"input":0.411,"output":4.11,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-vl-plus"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":512000},"cost":{"input":0.288,"output":1.152},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3.8-max-free":{"id":"qwen3.8-max-free","name":"Qwen3.8 Max (free)","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.8-max"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.6,"output":3,"cache_read":0.1},"canonical_model_id":"moonshotai/kimi-k2.5"},"alicloud-deepseek-v4-flash":{"id":"alicloud-deepseek-v4-flash","name":"DeepSeek V4 Flash (Alibaba Cloud)","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"qwen3.8-2.4t-a95b":{"id":"qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":12.5,"output":75,"cache_read":1.25},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash-image"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"claude-sonnet-5-5":{"id":"claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":0,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.499,"cache_read":0.03,"input_audio":0.9999},"canonical_model_id":"google/gemini-2.5-flash"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.1,"output":5.5,"cache_read":0.11,"cache_write":1.375},"canonical_model_id":"anthropic/claude-haiku-4-5"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":64000},"cost":{"input":1.08,"output":3.0888,"cache_read":0.054},"canonical_model_id":"stepfun/step-5-preview"},"qwen3.8-flash":{"id":"qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1126,"output":0.380025,"cache_read":0.014075,"cache_write":0.1759375},"canonical_model_id":"alibaba/qwen3.8-flash"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.375,"output":4.675},"canonical_model_id":"meta/muse-spark-1.1"},"qwen3.8-omni-flash":{"id":"qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.1126,"output":0.380025,"cache_read":0.014075,"cache_write":0.1759375},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"alicloud-deepseek-v4-pro":{"id":"alicloud-deepseek-v4-pro","name":"DeepSeek V4 Pro (Alibaba Cloud)","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":1.69,"output":3.38,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4-pro"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"mimo-v2.5":{"id":"mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.155,"output":0.31,"cache_read":0.0031},"canonical_model_id":"xiaomi/mimo-v2.5"},"coding-glm-5.3-flash-free":{"id":"coding-glm-5.3-flash-free","name":"GLM-5.3-Flash (Coding, free)","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"nemotron-3.5-content-safety-free":{"id":"nemotron-3.5-content-safety-free","name":"Nemotron 3.5 Content Safety (free)","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nemotron-3-nano-omni-30b-a3b-reasoning-free":{"id":"nemotron-3-nano-omni-30b-a3b-reasoning-free","name":"Nemotron 3 Nano Omni 30B A3B Reasoning (free)","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"claude-fable-5-1":{"id":"claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":11,"output":55,"cache_read":0.275,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5-1"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2,"tiers":[{"input":2.5,"output":5,"cache_read":0.4,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":5,"cache_read":0.4}},"canonical_model_id":"xai/grok-4.3"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":false,"interleaved":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":120},"canonical_model_id":"google/gemini-3-pro-image"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"experimental":{"modes":{"fast":{"cost":{"input":1.5,"output":9,"cache_read":0.15},"provider":{"body":{"service_tier":"priority"}}}}},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"qwen3.6-plus":{"id":"qwen3.6-plus","name":"Qwen3.6 Plus","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":65536},"cost":{"input":0.282,"output":1.692,"cache_read":0.0282,"cache_write":0.3525,"tiers":[{"input":1.128,"output":6.768,"cache_read":0.1128,"cache_write":1.41,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.128,"output":6.768,"cache_read":0.1128,"cache_write":1.41}},"canonical_model_id":"alibaba/qwen3.6-plus"},"gpt-5.5-pro":{"id":"gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.5-pro"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"longcat-2.0":{"id":"longcat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.7746,"output":3.0984,"cache_read":0.015492},"canonical_model_id":"meituan/longcat-2.0"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"muse-spark-1.3":{"id":"muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.375,"output":4.675,"cache_read":0.165},"canonical_model_id":"meta/muse-spark-1.3"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"claude-opus-4-5-think":{"id":"claude-opus-4-5-think","name":"Claude Opus 4.5 Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.6918,"output":2.0754,"cache_read":0.023057694},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"coding-minimax-m2.7-highspeed":{"id":"coding-minimax-m2.7-highspeed","name":"Coding MiniMax M2.7 Highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128100},"cost":{"input":0.2,"output":0.2},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"coding-glm-5.1-free":{"id":"coding-glm-5.1-free","name":"Coding GLM 5.1 (free)","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.1"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax M2.7","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06,"cache_write":0.375},"canonical_model_id":"minimax/MiniMax-M2.7"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.142,"output":0.284,"cache_read":0.0284},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"qwen3.5-flash":{"id":"qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.0282,"output":0.282,"cache_read":0.00282,"cache_write":0.03525,"tiers":[{"input":0.1126,"output":1.126,"cache_read":0.01126,"cache_write":0.14075,"tier":{"type":"context","size":128000}},{"input":0.169,"output":1.69,"cache_read":0.0169,"cache_write":0.21125,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen3.5-plus":{"id":"qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.1096,"output":0.6576,"cache_read":0.01096,"cache_write":0.137,"tiers":[{"input":0.274,"output":1.644,"cache_read":0.0274,"cache_write":0.3425,"tier":{"type":"context","size":128000}},{"input":0.548,"output":3.288,"cache_read":0.0548,"cache_write":0.685,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1644,"output":0.9864,"tiers":[{"input":0.411,"output":2.466,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"mimo-v2.6-flash":{"id":"mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.155,"output":0.31,"cache_read":0.0031},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"glm-4.7-flash-free":{"id":"glm-4.7-flash-free","name":"GLM-4.7-Flash (free)","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"qwen3.5-122b-a10b":{"id":"qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.1126,"output":0.9008,"tiers":[{"input":0.2818,"output":2.2544,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"claude-opus-5-5":{"id":"claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"nemotron-3.5-lightning-free":{"id":"nemotron-3.5-lightning-free","name":"Nemotron 3.5 Lightning 30B A3B (free)","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_details"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"qwen3.6-flash":{"id":"qwen3.6-flash","name":"Qwen3.6 Flash","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":65536},"cost":{"input":0.169,"output":1.014,"cache_read":0.0169,"cache_write":0.21125,"tiers":[{"input":0.676,"output":4.056,"cache_read":0.0676,"cache_write":0.845,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":0.676,"output":4.056,"cache_read":0.0676,"cache_write":0.845}},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0.4508,"output":2.7048,"cache_read":0.09016,"cache_write":0.5635,"tiers":[{"input":0.902,"output":5.412,"cache_read":0.1804,"cache_write":1.1275,"tier":{"type":"context","size":32000}},{"input":1.3522,"output":8.1132,"cache_read":0.27044,"cache_write":1.69025,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3-max"},"ling-3.0-flash-free":{"id":"ling-3.0-flash-free","name":"Ling 3.0 Flash (free)","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"inclusionai/ling-3.0-flash"},"claude-sonnet-4-6-think":{"id":"claude-sonnet-4-6-think","name":"Claude Sonnet 4.6 Thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"glm-5.3":{"id":"glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.1268,"output":3.9438,"cache_read":0.2817},"canonical_model_id":"zhipuai/glm-5.3"},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"claude-opus-4-6-think":{"id":"claude-opus-4-6-think","name":"Claude Opus 4.6 Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":131072},"cost":{"input":1.69,"output":5.07,"cache_read":0.169,"cache_write":2.1125},"canonical_model_id":"alibaba/qwen3.7-max"},"glm-4.6v":{"id":"glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.137,"output":0.411,"cache_read":0.0274,"tiers":[{"input":0.274,"output":0.822,"cache_read":0.0548,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-4.6v"},"grok-build-0.1":{"id":"grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0.1562,"output":0.6248,"cache_read":0.03905},"canonical_model_id":"tencent/hy3"},"mimo-v2.6-pro-ultraspeed":{"id":"mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":4.8,"output":9.6,"cache_read":0.0384},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gpt-5.1-codex-mini":{"id":"gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-image-2.5-sunburst":{"id":"gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2.5-sunburst"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0.48,"output":0.96,"cache_read":0.00384},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"doubao-seed-2-0-pro":{"id":"doubao-seed-2-0-pro","name":"Doubao Seed 2.0 Pro","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4958,"output":2.479,"cache_read":0.09916,"tiers":[{"input":0.7436,"output":3.718,"cache_read":0.14872,"tier":{"type":"context","size":32000}},{"input":1.4872,"output":7.436,"cache_read":0.29744,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"deep-deepseek-v4-pro":{"id":"deep-deepseek-v4-pro","name":"DeepSeek V4 Pro (DeepSeek)","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.478,"output":0.956,"cache_read":0.004302},"canonical_model_id":"deepseek/deepseek-v4-pro"},"xiaomi-mimo-v2.5-free":{"id":"xiaomi-mimo-v2.5-free","name":"Xiaomi MiMo-V2.5 (free)","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"claude-opus-4-5":{"id":"claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]},{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"deepseek-v4-flash-0731-fast":{"id":"deepseek-v4-flash-0731-fast","name":"DeepSeek V4 Flash 0731 Fast","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.28,"output":1.4,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"coding-minimax-m2.7-free":{"id":"coding-minimax-m2.7-free","name":"Coding MiniMax M2.7 (Free)","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":128100},"cost":{"input":0,"output":0},"canonical_model_id":"minimax/MiniMax-M2.7"},"qwen3.6-35b-a3b":{"id":"qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.254,"output":1.524},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"output":131072},"cost":{"input":1.69,"output":5.07,"cache_read":0.169,"cache_write":2.1125},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-image-2":{"id":"gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":30,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-2"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.11268,"output":0.39438,"cache_read":0.02817},"canonical_model_id":"zhipuai/glm-5.3-flash"},"xiaomi-mimo-v2.5-pro":{"id":"xiaomi-mimo-v2.5-pro","name":"Xiaomi MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":1.1,"output":3.3,"cache_read":0.22,"tiers":[{"input":2.2,"output":6.6,"cache_read":0.44,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2.2,"output":6.6,"cache_read":0.44}},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"coding-glm-5.2-free":{"id":"coding-glm-5.2-free","name":"GLM-5.2 (Coding, free)","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.2"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.499999,"cache_read":0.03,"input_audio":0.3},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"xiaomi-mimo-v2.5-pro-free":{"id":"xiaomi-mimo-v2.5-pro-free","name":"Xiaomi MiMo-V2.5-Pro (free)","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo-v2.5-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"status":"deprecated","cost":{"input":0,"output":0,"cache_read":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"qwen3.6-max-preview":{"id":"qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":240000,"output":64000},"status":"deprecated","cost":{"input":1.27,"output":7.61,"cache_read":0.1268,"cache_write":1.585,"tiers":[{"input":2.11,"output":12.67,"cache_read":0.2112,"cache_write":2.64,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"qwen3.5-35b-a3b":{"id":"qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.0564,"output":0.4512,"tiers":[{"input":0.2254,"output":1.8032,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.375,"output":4.675},"canonical_model_id":"meta/muse-spark-1.2"},"qwen3.5-27b":{"id":"qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.0846,"output":0.6768,"tiers":[{"input":0.2536,"output":2.0288,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":1000000},"cost":{"input":2.2,"output":6.6,"cache_read":0.55,"tiers":[{"input":4.4,"output":13.2,"cache_read":1.1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4.4,"output":13.2,"cache_read":1.1}},"canonical_model_id":"xai/grok-4.5"},"qwen3.8-flash-free":{"id":"qwen3.8-flash-free","name":"Qwen3.8 Flash (free)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]},{"type":"budget_tokens"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.8-flash"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":512,"max":24576}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"doubao-seed-2-0-mini-260428":{"id":"doubao-seed-2-0-mini-260428","name":"Doubao Seed 2.0 Mini 260428","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.030136,"output":0.30136,"cache_read":0.0060272,"input_audio":0.423,"tiers":[{"input":0.060274,"output":0.60274,"cache_read":0.0120548,"tier":{"type":"context","size":32000}},{"input":0.120548,"output":1.20548,"cache_read":0.0241096,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"kimi-k2.7-code-highspeed":{"id":"kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":1.9,"output":7.999,"cache_read":0.32167},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"qwen3.7-flash":{"id":"qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","max":262144}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"input":991808,"output":131072},"cost":{"input":0.0282,"output":0.1128,"cache_read":0.00564,"cache_write":0.03525,"tiers":[{"input":0.0845,"output":0.338,"cache_read":0.0169,"cache_write":0.105625,"tier":{"type":"context","size":32000}},{"input":0.169,"output":0.676,"cache_read":0.0338,"cache_write":0.21125,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"deepseek-v4-flash-vision-exp":{"id":"deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.155,"output":0.62,"cache_read":0.0031},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"coding-xiaomi-mimo-v2.5":{"id":"coding-xiaomi-mimo-v2.5","name":"Coding Xiaomi MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo-v2.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-05-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.08,"output":0.16,"cache_read":0.0016},"canonical_model_id":"xiaomi/mimo-v2.5"}}},"pioneer":{"id":"pioneer","env":["PIONEER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.fastino.ai/v1","name":"Pioneer","doc":"https://agent.pioneer.ai/llms.txt","models":{"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"output":128000},"status":"deprecated","cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":5},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-haiku-4-5":{"id":"claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":11,"output":55,"cache_read":1.1,"cache_write":13.75},"canonical_model_id":"anthropic/claude-fable-5"},"claude-opus-5-fast":{"id":"claude-opus-5-fast","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":2.5,"output":15,"cache_read":0.25,"cache_write":3.125},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":1,"output":6,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"openai/gpt-5.6-luna"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":5,"output":30,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"openai/gpt-5.6-sol"},"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16":{"id":"nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":4096},"cost":{"input":0.07,"output":0.2,"cache_read":0.035,"cache_write":0.07},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4":{"id":"nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":4096},"cost":{"input":0.08,"output":0.45,"cache_read":0.04,"cache_write":0.08},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"fastino/GLiNER-2.5-Decide":{"id":"fastino/GLiNER-2.5-Decide","name":"GLiNER-2.5-Decide","description":"Classification-tuned GLiNER2 large. Entities, relations, and JSON extraction still work.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-09-24","last_updated":"2026-09-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-large-v1":{"id":"fastino/gliner2-large-v1","name":"GLiNER2 Large","description":"Higher-accuracy NER; broader entity coverage.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-base-v1":{"id":"fastino/gliner2-base-v1","name":"GLiNER2 Base","description":"Named entity recognition; zero-shot span extraction.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":2048,"output":2048},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-privacy-filter-PII-multi":{"id":"fastino/gliner2-privacy-filter-PII-multi","name":"GLiNER2 Privacy Filter PII (Multi)","description":"Multilingual PII entity detection across 40+ languages; non-trainable encoder.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2.5-multi-v1":{"id":"fastino/gliner2.5-multi-v1","name":"GLiNER 2.5 Multi","description":"Multilingual boundary NER and span extraction; non-trainable encoder.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-08-24","last_updated":"2026-08-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliguard-PII-multi":{"id":"fastino/gliguard-PII-multi","name":"GLiNER2-Guardrails-PII-Multi","description":"A 300M-parameter multilingual model that runs LLM safety moderation and PII detection in a single forward pass.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.15,"output":0.15,"cache_read":0.15,"cache_write":0.15}},"fastino/gliner2.5-base-v1":{"id":"fastino/gliner2.5-base-v1","name":"GLiNER 2.5 Base","description":"English boundary NER and structured extraction with LoRA fine-tuning.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-08-24","last_updated":"2026-08-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliner2-multi-v1":{"id":"fastino/gliner2-multi-v1","name":"GLiNER2 Multi","description":"Multilingual NER for cross-lingual extraction.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-11-30","last_updated":"2025-11-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"fastino/gliguard-LLMGuardrails-300M":{"id":"fastino/gliguard-LLMGuardrails-300M","name":"GLiGuard LLM Guardrails 300M","description":"LLM safety, PII, and adversarial-attack guardrails; 300M-param encoder.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.15,"output":0.15,"cache_read":0.15,"cache_write":0.15}},"fastino/gliner2-multi-large-v1":{"id":"fastino/gliner2-multi-large-v1","name":"GLiNER2 Multi Large","description":"High-accuracy multilingual NER and span extraction.","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-11-30","last_updated":"2025-11-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"cost":{"input":0.03,"output":0,"cache_read":0.03,"cache_write":0.03}},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K3-Fast":{"id":"moonshotai/Kimi-K3-Fast","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":4.5,"output":22.5,"cache_read":0.45,"cache_write":4.5},"canonical_model_id":"moonshotai/kimi-k3"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0.1,"output":0.2,"cache_read":0.0197,"cache_write":0.1},"canonical_model_id":"deepseek/deepseek-v4-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1040000,"output":128000},"status":"deprecated","cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":1.4},"canonical_model_id":"zhipuai/glm-5.2"}}},"digitalocean":{"id":"digitalocean","env":["DIGITALOCEAN_ACCESS_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.do-ai.run/v1","name":"DigitalOcean","doc":"https://docs.digitalocean.com/products/gradient-ai-platform/details/models/","models":{"openai-gpt-6-luna":{"id":"openai-gpt-6-luna","name":"OpenAI GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai-gpt-5.6-terra":{"id":"openai-gpt-5.6-terra","name":"OpenAI GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"nemotron-3-nano-omni":{"id":"nemotron-3-nano-omni","name":"Nemotron 3 Nano Omni","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.5,"output":0.9,"cache_read":0.1},"canonical_model_id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"},"openai-gpt-oss-20b":{"id":"openai-gpt-oss-20b","name":"OpenAI GPT-oss-20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.05,"output":0.45,"cache_read":0.01},"canonical_model_id":"openai/gpt-oss-20b"},"deepseek-3.2":{"id":"deepseek-3.2","name":"Deepseek 3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"cost":{"input":0.5,"output":1.6,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v3.2"},"openai-gpt-6-astra":{"id":"openai-gpt-6-astra","name":"OpenAI GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai-gpt-5-nano":{"id":"openai-gpt-5-nano","name":"OpenAI GPT-5 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.21},"canonical_model_id":"zhipuai/glm-5.2"},"openai-gpt-6-sol":{"id":"openai-gpt-6-sol","name":"OpenAI GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"deepseek-4-flash":{"id":"deepseek-4-flash","name":"Deepseek V4 Flash","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek-flash","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"minimax-m3.1-flash":{"id":"minimax-m3.1-flash","name":"MiniMax M3.1-Flash (Public Preview)","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"deepseek-v4.1-flash":{"id":"deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"bge-m3":{"id":"bge-m3","type":"embedding","name":"BGE M3","description":"Multilingual embedding model combining dense, sparse, and multi-vector retrieval","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-30","last_updated":"2024-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":1024},"cost":{"input":0.02,"output":0},"canonical_model_id":"baai/bge-m3"},"llama-4-maverick":{"id":"llama-4-maverick","name":"Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.25,"output":0.87,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"gemma-4-31B-it":{"id":"gemma-4-31B-it","name":"Gemma 4","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192},"cost":{"input":0.18,"output":0.5,"cache_read":0.036},"canonical_model_id":"google/gemma-4-31b-it"},"deepseek-v4.1-flash-1017":{"id":"deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","xhigh","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"anthropic-claude-opus-4.8":{"id":"anthropic-claude-opus-4.8","name":"Anthropic Claude Opus 4.8","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"Deepseek V4 Pro","description":"Flagship DeepSeek model for coding, reasoning, and agentic work","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"cost":{"input":1.74,"output":3.48,"cache_read":0.348},"canonical_model_id":"deepseek/deepseek-v4-pro"},"openai-gpt-5-mini":{"id":"openai-gpt-5-mini","name":"OpenAI GPT-5 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"multi-qa-mpnet-base-dot-v1":{"id":"multi-qa-mpnet-base-dot-v1","type":"embedding","name":"Multi-QA-mpnet-base-dot-v1","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2021-08-23","last_updated":"2021-08-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":768},"cost":{"input":0.009,"output":0},"canonical_model_id":"sentence-transformers/multi-qa-mpnet-base-dot-v1"},"anthropic-claude-fable-5.1":{"id":"anthropic-claude-fable-5.1","name":"Anthropic Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"openai-gpt-5":{"id":"openai-gpt-5","name":"OpenAI GPT-5","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"reflection-beam":{"id":"reflection-beam","name":"Beam","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.4,"output":1.2,"cache_read":0.04}},"openai-gpt-5.5":{"id":"openai-gpt-5.5","name":"OpenAI GPT-5.5","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai-gpt-4o-mini":{"id":"openai-gpt-4o-mini","name":"OpenAI GPT-4o mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"nemotron-3-ultra-550b":{"id":"nemotron-3-ultra-550b","name":"Nemotron 3 Ultra","description":"Flagship Nemotron model for high-throughput reasoning and complex agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.9,"output":1.7,"cache_read":0.18},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"mistral-3-14B":{"id":"mistral-3-14B","name":"Ministral 3 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":128000},"cost":{"input":0.2,"output":0.2,"cache_read":0.04},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"anthropic-claude-opus-5":{"id":"anthropic-claude-opus-5","name":"Anthropic Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic-claude-opus-4.7":{"id":"anthropic-claude-opus-4.7","name":"Anthropic Claude Opus 4.7","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"e5-large-v2":{"id":"e5-large-v2","type":"embedding","name":"E5 Large v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-05","last_updated":"2023-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":1024},"cost":{"input":0.02,"output":0},"canonical_model_id":"intfloat/e5-large-v2"},"anthropic-claude-5-sonnet":{"id":"anthropic-claude-5-sonnet","name":"Anthropic Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"all-mini-lm-l6-v2":{"id":"all-mini-lm-l6-v2","type":"embedding","name":"All-MiniLM-L6-v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2021-08-30","last_updated":"2021-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256,"output":384},"cost":{"input":0.009,"output":0},"canonical_model_id":"sentence-transformers/all-minilm-l6-v2"},"openai-gpt-5.6-sol":{"id":"openai-gpt-5.6-sol","name":"OpenAI GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic-claude-sonnet-5.5":{"id":"anthropic-claude-sonnet-5.5","name":"Anthropic Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"anthropic-claude-4.5-sonnet":{"id":"anthropic-claude-4.5-sonnet","name":"Anthropic Claude 4.5 Sonnet","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.6"},"anthropic-claude-fable-5":{"id":"anthropic-claude-fable-5","name":"Anthropic Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"openai-gpt-5.4-mini":{"id":"openai-gpt-5.4-mini","name":"OpenAI GPT-5.4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"stable-diffusion-3.5-large":{"id":"stable-diffusion-3.5-large","type":"image","name":"Stable Diffusion 3.5 Large","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"stable-diffusion","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-10-22","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":256,"output":1}},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"openai-gpt-5.6-luna":{"id":"openai-gpt-5.6-luna","name":"OpenAI GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"bge-reranker-v2-m3":{"id":"bge-reranker-v2-m3","type":"reranking","name":"BGE Reranker v2 M3","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-15","last_updated":"2024-03-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":1},"cost":{"input":0.01,"output":0},"canonical_model_id":"baai/bge-reranker-v2-m3"},"anthropic-claude-opus-4.5":{"id":"anthropic-claude-opus-4.5","name":"Anthropic Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-5"},"nemotron-nano-12b-v2-vl":{"id":"nemotron-nano-12b-v2-vl","name":"Nemotron-nano 12b v2-vl","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.2,"output":0.6,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-nano-12b-v2-vl"},"qwen3-embedding-0.6b":{"id":"qwen3-embedding-0.6b","type":"embedding","name":"Qwen3 Embedding 0.6B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-03","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":1024},"status":"beta","cost":{"input":0.04,"output":0}},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"qwen3.5-397b-a17b":{"id":"qwen3.5-397b-a17b","name":"Qwen 3.5 397B A17B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.55,"output":3.5,"cache_read":0.111},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"gte-large-en-v1.5":{"id":"gte-large-en-v1.5","type":"embedding","name":"GTE Large (v1.5)","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"text-embedding","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-27","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":1024},"cost":{"input":0.09,"output":0}},"openai-o3":{"id":"openai-o3","name":"OpenAI o3","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"glm-5.3":{"id":"glm-5.3","name":"GLM5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":128000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"openai-gpt-oss-120b":{"id":"openai-gpt-oss-120b","name":"OpenAI GPT-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.1,"output":0.7,"cache_read":0.02},"canonical_model_id":"openai/gpt-oss-120b"},"arcee-trinity-large-thinking":{"id":"arcee-trinity-large-thinking","name":"Arcee Trinity Large Thinking (Public Preview)","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32000},"cost":{"input":0.25,"output":0.9,"cache_read":0.06},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"openai-gpt-5.2-pro":{"id":"openai-gpt-5.2-pro","name":"OpenAI GPT-5.2 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":21,"output":168},"canonical_model_id":"openai/gpt-5.2-pro"},"openai-gpt-4.1":{"id":"openai-gpt-4.1","name":"OpenAI GPT-4.1","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"minimax-m2.5":{"id":"minimax-m2.5","name":"MiniMax M2.5 (Public Preview)","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax-m2.5","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08","release_date":"2026-02-12","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"cost":{"input":0.3,"output":1.2,"cache_read":0.06}},"openai-gpt-5.3-codex":{"id":"openai-gpt-5.3-codex","name":"OpenAI GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"mimo-v2.5-pro":{"id":"mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.8,"output":3,"cache_read":0.16},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"openai-gpt-5.2":{"id":"openai-gpt-5.2","name":"OpenAI GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai-gpt-5.4-pro":{"id":"openai-gpt-5.4-pro","name":"OpenAI GPT-5.4 Pro","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","high","xhigh"]}],"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":30,"output":180,"tiers":[{"input":60,"output":270,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":60,"output":270}},"canonical_model_id":"openai/gpt-5.4-pro"},"anthropic-claude-opus-4.6":{"id":"anthropic-claude-opus-4.6","name":"Anthropic Claude Opus 4.6","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25,"tiers":[{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":10,"output":37.5,"cache_read":1,"cache_write":12.5}},"canonical_model_id":"anthropic/claude-opus-4-6"},"openai-gpt-5.4-nano":{"id":"openai-gpt-5.4-nano","name":"OpenAI GPT-5.4 Nano","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"anthropic-claude-haiku-4.5":{"id":"anthropic-claude-haiku-4.5","name":"Anthropic Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic-claude-4.6-sonnet":{"id":"anthropic-claude-4.6-sonnet","name":"Anthropic Claude Sonnet 4.6","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75,"tiers":[{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":6,"output":22.5,"cache_read":0.6,"cache_write":7.5}},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8-Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":262144},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"alibaba/qwen3.8-max"},"glm-5.3-flash":{"id":"glm-5.3-flash","name":"GLM5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"openai-gpt-6-1-sol":{"id":"openai-gpt-6-1-sol","name":"OpenAI GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai-gpt-image-2":{"id":"openai-gpt-image-2","type":"image","name":"OpenAI GPT Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image","text"]},"open_weights":false,"limit":{"context":0,"output":16384},"cost":{"input":8,"output":30},"canonical_model_id":"openai/gpt-image-2"},"wan2-2-t2v-a14b":{"id":"wan2-2-t2v-a14b","type":"video","name":"Wan2.2-T2V-A14B","description":"Video model for prompt-guided generation, editing, and motion workflows","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-07-28","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["video"]},"open_weights":true,"limit":{"context":100,"output":1}},"openai-gpt-4o":{"id":"openai-gpt-4o","name":"OpenAI GPT-4o","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai-o3-mini":{"id":"openai-o3-mini","name":"OpenAI o3 mini","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"anthropic-claude-5.5-haiku":{"id":"anthropic-claude-5.5-haiku","name":"Anthropic Claude 5.5 Haiku","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic-claude-opus-5.5":{"id":"anthropic-claude-opus-5.5","name":"Anthropic Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.4,"cache_write":10,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.4,"cache_write":10}},"canonical_model_id":"anthropic/claude-opus-5-5"},"openai-gpt-image-1":{"id":"openai-gpt-image-1","type":"image","name":"GPT Image 1","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0},"cost":{"input":5,"output":40,"cache_read":1.25},"canonical_model_id":"openai/gpt-image-1"},"openai-gpt-image-1.5":{"id":"openai-gpt-image-1.5","type":"image","name":"OpenAI GPT Image 1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":16384},"cost":{"input":5,"output":10,"cache_read":1},"canonical_model_id":"openai/gpt-image-1.5"},"nemotron-3.5-super-vl":{"id":"nemotron-3.5-super-vl","name":"NVIDIA Nemotron 3.5 Super VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":0.65,"cache_read":0.06},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"openai-gpt-5.4":{"id":"openai-gpt-5.4","name":"OpenAI GPT-5.4","description":"Frontier GPT model for professional reasoning, coding, and multimodal work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"qwen3-tts-voicedesign":{"id":"qwen3-tts-voicedesign","type":"speech","name":"Qwen3 TTS VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-30","modalities":{"input":["text"],"output":["audio"]},"open_weights":true,"limit":{"context":32768,"output":1},"provider":{"npm":"@ai-sdk/openai"}},"openai-o1":{"id":"openai-o1","name":"OpenAI o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"fal-ai/fast-sdxl":{"id":"fal-ai/fast-sdxl","type":"image","name":"Fast SDXL","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"stable-diffusion","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-07-26","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0}},"fal-ai/elevenlabs/tts/multilingual-v2":{"id":"fal-ai/elevenlabs/tts/multilingual-v2","type":"speech","name":"ElevenLabs Multilingual TTS v2","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"elevenlabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-08-22","last_updated":"2023-08-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0},"canonical_model_id":"elevenlabs/eleven_multilingual_v2"},"fal-ai/stable-audio-25/text-to-audio":{"id":"fal-ai/stable-audio-25/text-to-audio","type":"speech","name":"Stable Audio 2.5 (Text-to-Audio)","description":"Speech generation model for controllable voice, narration, and audio delivery","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-08","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":0,"output":0}},"fal-ai/flux/schnell":{"id":"fal-ai/flux/schnell","type":"image","name":"FLUX.1 [schnell]","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-01","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0}}}},"github-copilot":{"id":"github-copilot","env":["GITHUB_TOKEN"],"npm":"@ai-sdk/openai-compatible","api":"https://api.githubcopilot.com","name":"GitHub Copilot","doc":"https://docs.github.com/en/copilot","models":{"kimi-k2.7-code":{"id":"kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":224000,"output":32000},"status":"deprecated","cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"gemini-4-argon":{"id":"gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"google/gemini-4-argon"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]},{"type":"budget_tokens","min":256,"max":24000}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":128000,"output":64000},"status":"deprecated","cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"gpt-6-astra":{"id":"gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"claude-opus-4.7":{"id":"claude-opus-4.7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":168000,"output":32000},"status":"deprecated","experimental":{"modes":{"fast":{"cost":{"input":30,"output":150,"cache_read":3,"cache_write":37.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gemini-3.8-flash":{"id":"gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":264000,"input":128000,"output":64000},"status":"deprecated","cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"grok-4.7":{"id":"grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":372000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.7"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":372000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.6"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai"},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"gpt-6-luna":{"id":"gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"claude-haiku-4.5":{"id":"claude-haiku-4.5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":136000,"output":64000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"gpt-6-sol":{"id":"gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"claude-fable-5.1":{"id":"claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"mai-code-1.1-flash":{"id":"mai-code-1.1-flash","name":"MAI-Code-1.1-Flash","description":"Microsoft coding model with native vision support, optimized for fast and efficient software development","family":"mai","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":128000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"microsoft/mai-code-1.1-flash"},"claude-opus-5.5":{"id":"claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"gpt-6.1-sol":{"id":"gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]},{"type":"budget_tokens","min":256,"max":32000}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":936000,"output":64000},"status":"deprecated","cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"claude-haiku-5.5":{"id":"claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"claude-sonnet-5.5":{"id":"claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":372000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":2,"output":6,"cache_read":0.5,"tiers":[{"input":4,"output":12,"cache_read":1,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":4,"output":12,"cache_read":1}},"canonical_model_id":"xai/grok-4.5"},"claude-opus-4.8":{"id":"claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":168000,"output":64000},"experimental":{"modes":{"fast":{"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"provider":{"body":{"speed":"fast"},"headers":{"anthropic-beta":"fast-mode-2026-02-01"}}}}},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-8"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"claude-sonnet-4.6":{"id":"claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":32000}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":168000,"output":32000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.githubcopilot.com/v1"},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"}}},"siliconflow":{"id":"siliconflow","env":["SILICONFLOW_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.siliconflow.com/v1","name":"SiliconFlow","doc":"https://cloud.siliconflow.com/models","models":{"google/gemma-4-31B-it":{"id":"google/gemma-4-31B-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.75,"output":1,"cache_read":0.25},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemma-4-12B-it":{"id":"google/gemma-4-12B-it","name":"Gemma 4 12B IT","description":"Compact Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"google/gemma-4-12b-it"},"google/gemma-4-26B-A4B-it":{"id":"google/gemma-4-26B-A4B-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.14,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"ByteDance-Seed/Seed-OSS-36B-Instruct":{"id":"ByteDance-Seed/Seed-OSS-36B-Instruct","name":"ByteDance-Seed/Seed-OSS-36B-Instruct","description":"Tool-capable chat model for instruction following and agentic application workflows","family":"seed","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-04","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.21,"output":0.57}},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131000},"cost":{"input":0.3,"output":1.2,"cache_read":0.06},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.5":{"id":"MiniMaxAI/MiniMax-M2.5","name":"MiniMaxAI/MiniMax-M2.5","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":197000,"output":131000},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.5"},"tencent/Hunyuan-A13B-Instruct":{"id":"tencent/Hunyuan-A13B-Instruct","name":"tencent/Hunyuan-A13B-Instruct","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"hunyuan","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.57}},"tencent/Hy3":{"id":"tencent/Hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":192000,"output":262144},"cost":{"input":0.132,"output":0.528,"cache_read":0.033},"canonical_model_id":"tencent/hy3"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"moonshotai/Kimi-K2.6","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.77,"output":3.4,"cache_read":0.14},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.85916,"output":3.8,"cache_read":0.17993},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":262000},"cost":{"input":2.7,"output":13.5,"cache_read":0.27},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/Kimi-K2.5":{"id":"moonshotai/Kimi-K2.5","name":"moonshotai/Kimi-K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.45,"output":2.25,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"Qwen/Qwen3-8B":{"id":"Qwen/Qwen3-8B","name":"Qwen/Qwen3-8B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.06,"output":0.06},"canonical_model_id":"alibaba/qwen3-8b"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen/Qwen3-VL-30B-A3B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.29,"output":1},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3.5-27B":{"id":"Qwen/Qwen3.5-27B","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":2},"canonical_model_id":"alibaba/qwen3.5-27b"},"Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen/Qwen3-Coder-480B-A35B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.25,"output":1},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3-VL-32B-Thinking":{"id":"Qwen/Qwen3-VL-32B-Thinking","name":"Qwen/Qwen3-VL-32B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.2,"output":1.5}},"Qwen/Qwen3-VL-8B-Instruct":{"id":"Qwen/Qwen3-VL-8B-Instruct","name":"Qwen/Qwen3-VL-8B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.18,"output":0.68}},"Qwen/Qwen3-14B":{"id":"Qwen/Qwen3-14B","name":"Qwen/Qwen3-14B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-04-30","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.07,"output":0.28}},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen/Qwen3-32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.57},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":3.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3.8-2.4T-A95B":{"id":"Qwen/Qwen3.8-2.4T-A95B","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":131000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-2.4t-a95b"},"Qwen/Qwen3-VL-30B-A3B-Thinking":{"id":"Qwen/Qwen3-VL-30B-A3B-Thinking","name":"Qwen/Qwen3-VL-30B-A3B-Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-11","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.29,"output":1}},"Qwen/Qwen3-Coder-30B-A3B-Instruct":{"id":"Qwen/Qwen3-Coder-30B-A3B-Instruct","name":"Qwen/Qwen3-Coder-30B-A3B-Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.07,"output":0.28},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"Qwen/Qwen3-30B-A3B-Instruct-2507":{"id":"Qwen/Qwen3-30B-A3B-Instruct-2507","name":"Qwen/Qwen3-30B-A3B-Instruct-2507","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.09,"output":0.3},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"Qwen/Qwen3.5-35B-A3B":{"id":"Qwen/Qwen3.5-35B-A3B","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.24,"output":1.8,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"Qwen/Qwen3.6-35B-A3B":{"id":"Qwen/Qwen3.6-35B-A3B","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.24,"output":1.8,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-VL-32B-Instruct":{"id":"Qwen/Qwen3-VL-32B-Instruct","name":"Qwen/Qwen3-VL-32B-Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-21","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":262000},"cost":{"input":0.2,"output":0.6}},"Qwen/Qwen2.5-7B-Instruct":{"id":"Qwen/Qwen2.5-7B-Instruct","name":"Qwen/Qwen2.5-7B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0.05,"output":0.05},"canonical_model_id":"alibaba/qwen2.5-7b-instruct"},"Qwen/Qwen2.5-72B-Instruct":{"id":"Qwen/Qwen2.5-72B-Instruct","name":"Qwen/Qwen2.5-72B-Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2024-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":33000,"output":4000},"cost":{"input":0.59,"output":0.59}},"Qwen/Qwen3.5-9B":{"id":"Qwen/Qwen3.5-9B","name":"Qwen/Qwen3.5-9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.15},"canonical_model_id":"alibaba/qwen3.5-9b"},"Qwen/Qwen3.5-122B-A10B":{"id":"Qwen/Qwen3.5-122B-A10B","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.26,"output":2.08},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"Qwen/Qwen3.5-397B-A17B":{"id":"Qwen/Qwen3.5-397B-A17B","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"status":"deprecated","cost":{"input":0.39,"output":2.34},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"deepseek-ai/DeepSeek-V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp":{"id":"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.44,"output":1.32,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"deepseek-ai/DeepSeek-V3.2-Exp":{"id":"deepseek-ai/DeepSeek-V3.2-Exp","name":"deepseek-ai/DeepSeek-V3.2-Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-10","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":0.41}},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"deepseek-ai/DeepSeek-R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.5,"output":2.18},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.50162,"output":3.135,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.13,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3":{"id":"deepseek-ai/DeepSeek-V3","name":"deepseek-ai/DeepSeek-V3","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.25,"output":1},"canonical_model_id":"deepseek/deepseek-v3-0324"},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"deepseek-ai/DeepSeek-V3.1-Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-29","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":1}},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"deepseek-ai/DeepSeek-V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"output":164000},"cost":{"input":0.27,"output":0.42,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.22,"output":0.66,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek-ai/DeepSeek-V4-Pro-0813":{"id":"deepseek-ai/DeepSeek-V4-Pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"meituan-longcat/LongCat-2.0":{"id":"meituan-longcat/LongCat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":131072},"cost":{"input":0.75,"output":2.95,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"openai/gpt-oss-120b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":8000},"cost":{"input":0.05,"output":0.45},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"openai/gpt-oss-20b","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":8000},"cost":{"input":0.04,"output":0.18},"canonical_model_id":"openai/gpt-oss-20b"},"stepfun-ai/Step-3.5-Flash":{"id":"stepfun-ai/Step-3.5-Flash","name":"stepfun-ai/Step-3.5-Flash","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"output":66000},"cost":{"input":0.1,"output":0.3},"canonical_model_id":"stepfun/step-3.5-flash"},"zai-org/GLM-4.5-Air":{"id":"zai-org/GLM-4.5-Air","name":"zai-org/GLM-4.5-Air","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.86},"canonical_model_id":"zhipuai/glm-4.5-air"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"zai-org/GLM-5","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131072},"cost":{"input":0.95,"output":2.55,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"zai-org/GLM-5.1","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":205000,"output":131000},"cost":{"input":1.19,"output":3.74,"cache_read":0.6,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":0.15,"output":0.5,"cache_read":0.03,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1049000,"output":262000},"cost":{"input":1.302,"output":4.092,"cache_read":0.26,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"},"zai-org/GLM-5V-Turbo":{"id":"zai-org/GLM-5V-Turbo","name":"zai-org/GLM-5V-Turbo","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24,"cache_write":0},"canonical_model_id":"zhipuai/glm-5v-turbo"},"nex-agi/Nex-N2-Pro":{"id":"nex-agi/Nex-N2-Pro","name":"Nex-N2-Pro","description":"Open agentic MoE model (397B total, 17B active) for coding, tool use, and research workflows","attachment":true,"reasoning":true,"reasoning_options":[{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"status":"deprecated","cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"nex-agi/nex-n2-pro"},"inclusionAI/Ling-flash-2.0":{"id":"inclusionAI/Ling-flash-2.0","name":"inclusionAI/Ling-flash-2.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-18","last_updated":"2025-11-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.14,"output":0.57}}}},"stackit":{"id":"stackit","env":["STACKIT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.openai-compat.model-serving.eu01.onstackit.cloud/v1","name":"STACKIT","doc":"https://docs.stackit.cloud/products/data-and-ai/ai-model-serving/basics/available-shared-models","models":{"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":37000,"output":4096},"status":"deprecated","cost":{"input":0.53,"output":0.76},"canonical_model_id":"google/gemma-3-27b-it"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"status":"deprecated","cost":{"input":0.53,"output":0.76},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen3-VL-Embedding-8B":{"id":"Qwen/Qwen3-VL-Embedding-8B","type":"embedding","name":"Qwen3-VL Embedding 8B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":4096},"cost":{"input":0.09,"output":0}},"Qwen/Qwen3-VL-235B-A22B-Instruct-FP8":{"id":"Qwen/Qwen3-VL-235B-A22B-Instruct-FP8","name":"Qwen3-VL 235B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":218000,"output":16384},"cost":{"input":1.76,"output":2.05},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":8192},"cost":{"input":0.53,"output":0.76},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.18,"output":0.29},"canonical_model_id":"openai/gpt-oss-20b"},"intfloat/e5-mistral-7b-instruct":{"id":"intfloat/e5-mistral-7b-instruct","type":"embedding","name":"E5 Mistral 7B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"cost":{"input":0.02243,"output":0}},"cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic":{"id":"cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0.53,"output":0.76},"canonical_model_id":"meta/llama-3.3-70b-instruct"}}},"aixy":{"id":"aixy","env":["AIXY_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.aixy-gateway.com/v1","name":"Aixy","doc":"https://docs.aixy-gateway.com/integrations/overview","models":{"gemini/gemini-4-argon":{"id":"gemini/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"cost":{"input":2,"output":10,"cache_read":0.1,"input_audio":2},"canonical_model_id":"google/gemini-4-argon"},"gemini/gemini-3.8-flash":{"id":"gemini/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"input_audio":0.75},"canonical_model_id":"google/gemini-3.8-flash"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5,"tiers":[{"input":20,"output":75,"cache_read":2,"cache_write":25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":20,"output":75,"cache_read":2,"cache_write":25}},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.2,"output":0.75,"cache_read":0.02,"cache_write":0.25}},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5,"tiers":[{"input":4,"output":18,"cache_read":0.4,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":18,"cache_read":0.4,"cache_write":5}},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5,"tiers":[{"input":4,"output":15,"cache_read":0.2,"cache_write":5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":4,"output":15,"cache_read":0.2,"cache_write":5}},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25,"tiers":[{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":0.4,"output":1.8,"cache_read":0.04,"cache_write":0.5}},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"provider":{"npm":"@ai-sdk/openai","api":"https://api.aixy-gateway.com/v1","shape":"responses"},"cost":{"input":4,"output":20,"cache_read":0.4,"cache_write":5,"tiers":[{"input":8,"output":30,"cache_read":0.8,"cache_write":10,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":8,"output":30,"cache_read":0.8,"cache_write":10}},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"provider":{"npm":"@ai-sdk/anthropic","api":"https://api.aixy-gateway.com/v1"},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"}}},"unorouter":{"id":"unorouter","env":["UNOROUTER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.unorouter.com/v1","name":"UnoRouter","doc":"https://unorouter.com/models","models":{"deepseek-v4-flash:free":{"id":"deepseek-v4-flash:free","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"},"glm-5.2":{"id":"glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":1.6001,"output":5.0288},"canonical_model_id":"zhipuai/glm-5.2"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1857,"output":1.1142},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":1.8,"output":10.8},"canonical_model_id":"openai/gpt-5.4"},"gpt-5.4:free":{"id":"gpt-5.4:free","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.4"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.8999,"output":1.7999},"canonical_model_id":"deepseek/deepseek-v4-pro"},"glm-5.2:free":{"id":"glm-5.2:free","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-5.2"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":1.44,"output":7.2},"canonical_model_id":"anthropic/claude-sonnet-5"},"qwen3.5-397b-a17b:free":{"id":"qwen3.5-397b-a17b:free","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1875,"output":1.125},"canonical_model_id":"openai/gpt-5.5"},"nemotron-3-ultra-550b-a55b:free":{"id":"nemotron-3-ultra-550b-a55b:free","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"cost":{"input":0,"output":0},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"step-3.7-flash:free":{"id":"step-3.7-flash:free","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"stepfun/step-3.7-flash"},"gpt-5.5:free":{"id":"gpt-5.5:free","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-5.5"},"gemma-4-31b-it:free":{"id":"gemma-4-31b-it:free","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-4-31b-it"},"kimi-k2.6":{"id":"kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":1.2675,"output":5.3368},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax-m2.7":{"id":"minimax-m2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"status":"deprecated","cost":{"input":0.819,"output":3.276},"canonical_model_id":"minimax/MiniMax-M2.7"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.05,"output":8.4},"canonical_model_id":"openai/gpt-5.2"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"status":"deprecated","cost":{"input":0.425,"output":2.125},"canonical_model_id":"anthropic/claude-opus-4-8"},"glm-4.5-flash:free":{"id":"glm-4.5-flash:free","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1.2,"output":6},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"minimax-m2.7:free":{"id":"minimax-m2.7:free","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"minimax/MiniMax-M2.7"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0.0625,"output":0.125},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-v4-pro:free":{"id":"deepseek-v4-pro:free","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"status":"deprecated","cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-pro"}}},"cloudflare-workers-ai":{"id":"cloudflare-workers-ai","env":["CLOUDFLARE_ACCOUNT_ID","CLOUDFLARE_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/ai/v1","name":"Cloudflare Workers AI","doc":"https://developers.cloudflare.com/workers-ai/models/","models":{"@cf/mistralai/mistral-small-3.1-24b-instruct":{"id":"@cf/mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B Instruct","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"mistral/mistral-small-3-1-24b-instruct-2503"},"@cf/google/gemma-4-26b-a4b-it":{"id":"@cf/google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"@cf/nvidia/nemotron-3-120b-a12b":{"id":"@cf/nvidia/nemotron-3-120b-a12b","name":"Nemotron 3 Super 120B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high"]}],"tool_call":true,"interleaved":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"@cf/aisingapore/gemma-sea-lion-v4-27b-it":{"id":"@cf/aisingapore/gemma-sea-lion-v4-27b-it","name":"Gemma Sea Lion V4 27B It","description":"Gemma 3 27B tuned by AI Singapore for Southeast Asian languages and instruction following","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.351,"output":0.555},"canonical_model_id":"aisingapore/gemma-sea-lion-v4-27b-it"},"@cf/moonshotai/kimi-k2.7-code":{"id":"@cf/moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"@cf/moonshotai/kimi-k2.6":{"id":"@cf/moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"@cf/meta/llama-3.1-8b-instruct-fp8":{"id":"@cf/meta/llama-3.1-8b-instruct-fp8","name":"Llama 3.1 8B Instruct fp8","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000},"cost":{"input":0.152,"output":0.287},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"@cf/meta/llama-4-scout-17b-16e-instruct":{"id":"@cf/meta/llama-4-scout-17b-16e-instruct","name":"Llama 4 Scout 17B 16E Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":16384},"cost":{"input":0.27,"output":0.85},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"@cf/meta/llama-3.2-1b-instruct":{"id":"@cf/meta/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":60000,"output":60000},"cost":{"input":0.027,"output":0.201},"canonical_model_id":"meta/llama-3.2-1b"},"@cf/meta/llama-3.3-70b-instruct-fp8-fast":{"id":"@cf/meta/llama-3.3-70b-instruct-fp8-fast","name":"Llama 3.3 70B Instruct fp8 Fast","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":24000,"output":24000},"cost":{"input":0.293,"output":2.253},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"@cf/meta/llama-guard-3-8b":{"id":"@cf/meta/llama-guard-3-8b","name":"Llama Guard 3 8B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.484,"output":0.03},"canonical_model_id":"meta/llama-guard-3-8b"},"@cf/meta/llama-3.2-3b-instruct":{"id":"@cf/meta/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":80000},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"meta/llama-3.2-3b"},"@cf/meta/llama-3.2-11b-vision-instruct":{"id":"@cf/meta/llama-3.2-11b-vision-instruct","name":"Llama 3.2 11B Vision Instruct","description":"Open Llama multimodal model for image understanding and text reasoning","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"cost":{"input":0.0485,"output":0.676},"canonical_model_id":"meta/llama-3.2-11b-vision-instruct"},"@cf/ibm-granite/granite-4.0-h-micro":{"id":"@cf/ibm-granite/granite-4.0-h-micro","name":"Granite 4.0 H Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131000,"output":131000},"cost":{"input":0.017,"output":0.112},"canonical_model_id":"ibm/granite-4-h-micro"},"@cf/deepseek-ai/deepseek-v4.1-flash-1017":{"id":"@cf/deepseek-ai/deepseek-v4.1-flash-1017","name":"Deepseek V4.1 Flash 1017","description":"Fast DeepSeek model for efficient chat, coding help, and agent loops","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.3,"output":1.2,"cache_read":0.006}},"@cf/deepseek-ai/deepseek-r1-distill-qwen-32b":{"id":"@cf/deepseek-ai/deepseek-r1-distill-qwen-32b","name":"Deepseek R1 Distill Qwen 32B","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":80000,"output":80000},"cost":{"input":0.497,"output":4.881},"canonical_model_id":"deepseek/deepseek-r1"},"@cf/deepseek-ai/deepseek-v4-pro-0813":{"id":"@cf/deepseek-ai/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.32,"output":3.96,"cache_read":0.044},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"@cf/deepseek-ai/deepseek-v4-flash-0731":{"id":"@cf/deepseek-ai/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.44,"output":1.32,"cache_read":0.014},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"@cf/openai/gpt-oss-120b":{"id":"@cf/openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.35,"output":0.75},"canonical_model_id":"openai/gpt-oss-120b"},"@cf/openai/gpt-oss-20b":{"id":"@cf/openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open-weight GPT model for self-hosted reasoning and instruction-following workloads","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.2,"output":0.3},"canonical_model_id":"openai/gpt-oss-20b"},"@cf/qwen/qwen3-30b-a3b-fp8":{"id":"@cf/qwen/qwen3-30b-a3b-fp8","name":"Qwen3 30B A3b fp8","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.0509,"output":0.335},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"@cf/qwen/qwen2.5-coder-32b-instruct":{"id":"@cf/qwen/qwen2.5-coder-32b-instruct","name":"Qwen2.5 Coder 32B Instruct","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"@cf/qwen/qwq-32b":{"id":"@cf/qwen/qwq-32b","name":"Qwq 32B","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":24000,"output":24000},"cost":{"input":0.66,"output":1},"canonical_model_id":"alibaba/qwq-32b"},"@cf/qwen/qwen3.8-27b":{"id":"@cf/qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.45,"output":3.2,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.8-27b"},"@cf/zai-org/glm-4.7-flash":{"id":"@cf/zai-org/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"cost":{"input":0.0605,"output":0.4},"canonical_model_id":"zhipuai/glm-4.7-flash"},"@cf/zai-org/glm-5.2":{"id":"@cf/zai-org/glm-5.2","name":"Glm 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":256000},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"},"@cf/zai-org/glm-5.3":{"id":"@cf/zai-org/glm-5.3","name":"Glm 5.3","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"@cf/zai-org/glm-5.3-flash":{"id":"@cf/zai-org/glm-5.3-flash","name":"Glm 5.3 Flash","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"}}},"cohere":{"id":"cohere","env":["COHERE_API_KEY"],"npm":"@ai-sdk/cohere","name":"Cohere","doc":"https://docs.cohere.com/docs/models","models":{"tiny-aya-earth":{"id":"tiny-aya-earth","name":"Tiny Aya Earth","description":"Compact multilingual model specialized for West Asian and African languages","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-earth"},"tiny-aya-water":{"id":"tiny-aya-water","name":"Tiny Aya Water","description":"Compact multilingual model specialized for European and Asia-Pacific languages","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-water"},"command-a-translate-08-2025":{"id":"command-a-translate-08-2025","name":"Command A Translate","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-translate-08-2025"},"command-a-plus-05-2026":{"id":"command-a-plus-05-2026","name":"Command A Plus","description":"Cohere's stronger command model for multilingual agents and enterprise workflows","family":"command-a","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04-01","release_date":"2026-05-20","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-plus-05-2026"},"command-r7b-12-2024":{"id":"command-r7b-12-2024","name":"Command R7B","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-12-2024"},"command-r-plus-08-2024":{"id":"command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-r-plus-08-2024"},"command-a-reasoning-08-2025":{"id":"command-a-reasoning-08-2025","name":"Command A Reasoning","description":"Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows","family":"command-a","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-reasoning-08-2025"},"command-r-08-2024":{"id":"command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"cohere/command-r-08-2024"},"command-r7b-arabic-02-2025":{"id":"command-r7b-arabic-02-2025","name":"Command R7B Arabic","description":"Open Command R model optimized for Arabic enterprise chat, RAG, and cultural knowledge","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-02-27","last_updated":"2025-02-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.0375,"output":0.15},"canonical_model_id":"cohere/command-r7b-arabic-02-2025"},"tiny-aya-global":{"id":"tiny-aya-global","name":"Tiny Aya Global","description":"Compact multilingual model balanced across 70 languages and regions","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-global"},"c4ai-aya-vision-32b":{"id":"c4ai-aya-vision-32b","name":"Aya Vision 32B","description":"Open multilingual vision model for OCR, visual reasoning, and image question answering","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-04","last_updated":"2025-05-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4000},"canonical_model_id":"cohere/c4ai-aya-vision-32b"},"tiny-aya-fire":{"id":"tiny-aya-fire","name":"Tiny Aya Fire","description":"Compact multilingual model specialized for South Asian languages","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"canonical_model_id":"cohere/tiny-aya-fire"},"command-a-vision-07-2025":{"id":"command-a-vision-07-2025","name":"Command A Vision","description":"Cohere vision model for multilingual document analysis, OCR, and image understanding","family":"command-a","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-vision-07-2025"},"c4ai-aya-expanse-32b":{"id":"c4ai-aya-expanse-32b","name":"Aya Expanse 32B","description":"Open multilingual model optimized for generation across 23 languages","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-10-24","last_updated":"2024-10-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"cohere/c4ai-aya-expanse-32b"},"north-mini-code-1-0":{"id":"north-mini-code-1-0","name":"North Mini Code","description":"Cohere coding model for practical software engineering and agentic edits","family":"north","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":false,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"provider":{"npm":"@ai-sdk/openai-compatible","api":"https://api.cohere.ai/compatibility/v1"},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-mini-code-1-0"},"north-small-translate-09-2026":{"id":"north-small-translate-09-2026","name":"North Small Translate","description":"Open-weight translation model for more than 50 languages and locale variants","family":"north","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16000},"cost":{"input":0,"output":0},"canonical_model_id":"cohere/north-small-translate-1-0"},"command-a-03-2025":{"id":"command-a-03-2025","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8000},"cost":{"input":2.5,"output":10},"canonical_model_id":"cohere/command-a-03-2025"}}},"subconscious":{"id":"subconscious","env":["SUBCONSCIOUS_API_KEY"],"npm":"@ai-sdk/anthropic","api":"https://api.subconscious.dev/v1","name":"Subconscious","doc":"https://docs.subconscious.dev","models":{"subconscious/glm-5.3-marathon":{"id":"subconscious/glm-5.3-marathon","name":"GLM-5.3 Marathon","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":5000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"subconscious/deepseek-v4.1-flash-marathon":{"id":"subconscious/deepseek-v4.1-flash-marathon","name":"DeepSeek V4.1 Flash Marathon","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":5000000,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4.1-flash"}}},"modelis":{"id":"modelis","env":["MODELIS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://modelishub.com/v1","name":"Modelis","doc":"https://modelishub.com/pricing","models":{"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.435,"output":0.87},"canonical_model_id":"deepseek/deepseek-v4-pro"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["minimal","low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5},"canonical_model_id":"google/gemini-2.5-flash"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]},{"type":"budget_tokens"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10},"canonical_model_id":"google/gemini-2.5-pro"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.0983,"output":0.1966},"canonical_model_id":"deepseek/deepseek-v4-flash"},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":0.768,"output":3.072},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens"}],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":9},"canonical_model_id":"alibaba/qwen3.7-max"}}},"pendra":{"id":"pendra","env":["PENDRA_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.pendra.ai/api/v1","name":"Pendra","doc":"https://pendra.ai/docs/integrations/opencode","models":{"gpt-oss:120b":{"id":"gpt-oss:120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"openai/gpt-oss-120b"},"glm-4.7-flash":{"id":"glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"qwen3.6:27b":{"id":"qwen3.6:27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen3-coder:30b":{"id":"qwen3-coder:30b","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0,"output":0},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"llama3.3:70b":{"id":"llama3.3:70b","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"cost":{"input":0,"output":0},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"deepseek-v4-flash":{"id":"deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0,"output":0},"canonical_model_id":"deepseek/deepseek-v4-flash"}}},"empiriolabs":{"id":"empiriolabs","env":["EMPIRIOLABS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.empiriolabs.ai/v1","name":"EmpirioLabs AI","doc":"https://docs.empiriolabs.ai","models":{"qwen3-5-35b-a3b":{"id":"qwen3-5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.057,"output":0.459,"cache_read":0.057,"tiers":[{"input":0.229,"output":1.835,"cache_read":0.229,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen3-7-flash":{"id":"qwen3-7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":131072}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"tiers":[{"input":0.1,"output":0.4,"cache_read":0.02,"tier":{"type":"context","size":32000}},{"input":0.2,"output":0.8,"cache_read":0.04,"tier":{"type":"context","size":256000}}]},"canonical_model_id":"alibaba/qwen3.7-flash"},"muse-spark-1-1":{"id":"muse-spark-1-1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1},"canonical_model_id":"meta/muse-spark-1.1"},"qwen3-5-flash":{"id":"qwen3-5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":0.09,"output":0.368,"cache_read":0.09},"canonical_model_id":"alibaba/qwen3.5-flash"},"kimi-k2-6":{"id":"kimi-k2-6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16000},"cost":{"input":0.8939,"output":3.7131,"cache_read":0.1788},"canonical_model_id":"moonshotai/kimi-k2.6"},"minimax-m3-1-flash":{"id":"minimax-m3-1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":524288},"cost":{"input":0.1125,"output":0.45,"cache_read":0.01125},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"qwen3-6-flash":{"id":"qwen3-6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":64000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.25,"tiers":[{"input":1,"output":4,"cache_read":1,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1,"output":4,"cache_read":1}},"canonical_model_id":"alibaba/qwen3.6-flash"},"gemma-3-27b":{"id":"gemma-3-27b","name":"Gemma 3 27B","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"cost":{"input":0,"output":0},"canonical_model_id":"google/gemma-3-27b-it"},"qwen3-6-plus":{"id":"qwen3-6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.5,"tiers":[{"input":2,"output":6,"cache_read":2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":2,"output":6,"cache_read":2}},"canonical_model_id":"alibaba/qwen3.6-plus"},"seed-2-0-code":{"id":"seed-2-0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.4,"output":2.4,"cache_read":0.4,"tiers":[{"input":0.8,"output":4.8,"cache_read":0.8,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"fugu-max":{"id":"fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"muse-glimmer-30b":{"id":"muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.2,"output":0.8,"cache_read":0.05},"canonical_model_id":"meta/muse-glimmer-30b"},"deepseek-v4-pro":{"id":"deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.65,"output":3.3,"cache_read":1.65},"canonical_model_id":"deepseek/deepseek-v4-pro"},"glm-5-2":{"id":"glm-5-2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.2"},"glm-4-7-flash":{"id":"glm-4-7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.7-flash"},"qwen3-5-9b":{"id":"qwen3-5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.09,"output":0.13,"cache_read":0.045},"canonical_model_id":"alibaba/qwen3.5-9b"},"minimax-m2-7":{"id":"minimax-m2-7","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32768},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.7"},"qwen3-5-plus":{"id":"qwen3-5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1,"max":81920}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"status":"deprecated","cost":{"input":0.36,"output":2.21,"cache_read":0.36,"tiers":[{"input":1.08,"output":6.62,"cache_read":1.08,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.08,"output":6.62,"cache_read":1.08}},"canonical_model_id":"alibaba/qwen3.5-plus"},"minimax-m3":{"id":"minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":524288},"cost":{"input":0.225,"output":0.9,"cache_read":0.045,"tiers":[{"input":0.45,"output":1.8,"cache_read":0.09,"tier":{"type":"context","size":512000}}],"context_over_200k":{"input":0.45,"output":1.8,"cache_read":0.09}},"canonical_model_id":"minimax/MiniMax-M3"},"qwen3-8-27b":{"id":"qwen3-8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.17,"output":0.5,"cache_read":0.08},"canonical_model_id":"alibaba/qwen3.8-27b"},"step-3-7-flash":{"id":"step-3-7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":131072},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"step-5-preview":{"id":"step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":131072},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"kimi-k2-7-code":{"id":"kimi-k2-7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":131072},"cost":{"input":0.95,"output":4,"cache_read":0.95},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"minimax-m2-7-highspeed":{"id":"minimax-m2-7-highspeed","name":"MiniMax M2.7 Highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":32768},"status":"deprecated","cost":{"input":0.3,"output":1.2,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.7-highspeed"},"qwen3-8-flash":{"id":"qwen3-8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.16,"output":0.47,"cache_read":0.16},"canonical_model_id":"alibaba/qwen3.8-flash"},"seed-2-0-pro":{"id":"seed-2-0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.63,"output":3.79,"cache_read":0.63,"tiers":[{"input":1.26,"output":7.58,"cache_read":1.26,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"qwen3-6-35b-a3b":{"id":"qwen3-6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"cost":{"input":0.07,"output":0.42,"cache_read":0.035},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"fugu-ultra-v1-0":{"id":"fugu-ultra-v1-0","name":"Fugu Ultra v1.0","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":7.5,"output":45,"cache_read":1.5,"tiers":[{"input":15,"output":67.5,"cache_read":3,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":15,"output":67.5,"cache_read":3}},"canonical_model_id":"sakana/fugu-ultra"},"glm-5-1":{"id":"glm-5-1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":38912}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202000,"output":128000},"cost":{"input":0.825,"output":3.301,"cache_read":0.165,"tiers":[{"input":1.1,"output":3.851,"cache_read":0.22,"tier":{"type":"context","size":32000}}]},"canonical_model_id":"zhipuai/glm-5.1"},"kimi-k2-7-code-highspeed":{"id":"kimi-k2-7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":131072},"cost":{"input":1.9,"output":8,"cache_read":1.9},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"deepseek-v4-1-flash-1017":{"id":"deepseek-v4-1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.3,"output":1.2,"cache_read":0.3},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"qwen3-5-4b":{"id":"qwen3-5-4b","name":"Qwen3.5 4B","description":"Qwen3.5 4B is a low-cost multimodal reasoning model with 256K context, image and video input, function tools, and structured output.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1024,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.04,"output":0.07,"cache_read":0.02},"canonical_model_id":"alibaba/qwen3.5-4b"},"glm-4-5-flash":{"id":"glm-4-5-flash","name":"GLM 4.5 Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":98304},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.5-flash"},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":3,"output":15,"cache_read":3},"canonical_model_id":"moonshotai/kimi-k3"},"mimo-v2-6-flash":{"id":"mimo-v2-6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.14},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"deepseek-v4-pro-0813":{"id":"deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":1.32,"output":3.96,"cache_read":1.32},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek-v4-flash-0731":{"id":"deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":393216}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":393216},"cost":{"input":0.424,"output":1.272,"cache_read":0.424},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"fugu-ultra-v2-0":{"id":"fugu-ultra-v2-0","name":"Fugu Ultra v2.0","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra-v2"},"qwen3-5-397b-a17b":{"id":"qwen3-5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"status":"deprecated","cost":{"input":0.172,"output":1.032,"cache_read":0.172,"tiers":[{"input":0.43,"output":2.58,"cache_read":0.43,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen3-7-plus":{"id":"qwen3-7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":256000}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.4,"tiers":[{"input":1.2,"output":4.8,"cache_read":1.2,"tier":{"type":"context","size":256000}}],"context_over_200k":{"input":1.2,"output":4.8,"cache_read":1.2}},"canonical_model_id":"alibaba/qwen3.7-plus"},"glm-5-3-flash":{"id":"glm-5-3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.075,"output":0.25,"cache_read":0.075},"canonical_model_id":"zhipuai/glm-5.3-flash"},"mimo-v2-6-pro-ultraspeed":{"id":"mimo-v2-6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":4.35},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"qwen3-8-max-0902":{"id":"qwen3-8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":2},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"step-3-5-flash":{"id":"step-3-5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash"},"glm-5-3":{"id":"glm-5-3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":1.4},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3-7-max":{"id":"qwen3-7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":64000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536},"cost":{"input":2.5,"output":7.5,"cache_read":2.5},"canonical_model_id":"alibaba/qwen3.7-max"},"seed-2-0-lite":{"id":"seed-2-0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.31,"output":2.5,"cache_read":0.31,"tiers":[{"input":0.62,"output":5,"cache_read":0.62,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"muse-spark-1-3":{"id":"muse-spark-1-3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1},"canonical_model_id":"meta/muse-spark-1.3"},"qwen3-8-max":{"id":"qwen3-8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":2},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen3-5-27b":{"id":"qwen3-5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.086,"output":0.688,"cache_read":0.086,"tiers":[{"input":0.258,"output":2.064,"cache_read":0.258,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-27b"},"gemma-4-26b-a4b":{"id":"gemma-4-26b-a4b","name":"Gemma 4 26B-A4B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":128,"max":32768}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.05,"output":0.29,"cache_read":0.025},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"qwen3-8-omni-flash":{"id":"qwen3-8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":262144}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":0.94,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"muse-spark-1-2":{"id":"muse-spark-1-2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":1},"canonical_model_id":"meta/muse-spark-1.2"},"seed-2-0-mini":{"id":"seed-2-0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"cost":{"input":0.12,"output":0.5,"cache_read":0.12,"tiers":[{"input":0.24,"output":1,"cache_read":0.24,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"seed-2-1-turbo":{"id":"seed-2-1-turbo","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":65536},"cost":{"input":0.63,"output":3.13,"cache_read":0.63},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"mimo-v2-6-pro":{"id":"mimo-v2-6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.435},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"fugu-ultra-v1-1":{"id":"fugu-ultra-v1-1","name":"Fugu Ultra v1.1","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"sakana/fugu-ultra"},"qwen3-5-122b-a10b":{"id":"qwen3-5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.115,"output":0.917,"cache_read":0.115,"tiers":[{"input":0.287,"output":2.294,"cache_read":0.287,"tier":{"type":"context","size":128000}}]},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"mistral-small-4":{"id":"mistral-small-4","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.15},"canonical_model_id":"mistral/mistral-small-2603"},"step-3-5-flash-2603":{"id":"step-3-5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.02},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"glm-4-6v-flash":{"id":"glm-4-6v-flash","name":"GLM 4.6V Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0,"output":0},"canonical_model_id":"zhipuai/glm-4.6v-flash"},"qwen3-6-27b":{"id":"qwen3-6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]},{"type":"budget_tokens","min":1,"max":80000}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"cost":{"input":0.412564,"output":2.475384,"cache_read":0.412564},"canonical_model_id":"alibaba/qwen3.6-27b"}}},"aki-io":{"id":"aki-io","env":["AKI_IO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://aki.io/v1","name":"AKI.IO","doc":"https://aki.io/docs/","models":{"gpt-oss-120b":{"id":"gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"cost":{"input":0.168225,"output":0.616825},"canonical_model_id":"openai/gpt-oss-120b"},"gemma4-26b":{"id":"gemma4-26b","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"cost":{"input":0.11215,"output":0.56075},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"mistral4-119b":{"id":"mistral4-119b","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"cost":{"input":0.2243,"output":0.6729},"canonical_model_id":"mistral/mistral-small-2603"},"glm5.3-754b":{"id":"glm5.3-754b","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":81920},"cost":{"input":1.1215,"output":3.92525,"cache_read":0.280375},"canonical_model_id":"zhipuai/glm-5.3"},"qwen3.6-35b":{"id":"qwen3.6-35b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768},"status":"deprecated","cost":{"input":0.168225,"output":0.56075},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen3.8-27b":{"id":"qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.33645,"output":2.4673,"cache_read":0.11215},"canonical_model_id":"alibaba/qwen3.8-27b"},"deepseek-v4.1-flash-552b":{"id":"deepseek-v4.1-flash-552b","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":81920},"cost":{"input":0.280375,"output":1.1215,"cache_read":0.11215},"canonical_model_id":"deepseek/deepseek-v4.1-flash"}}},"qihang-ai":{"id":"qihang-ai","env":["QIHANG_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.qhaigc.net/v1","name":"QiHang","doc":"https://www.qhaigc.net/docs","models":{"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.07,"output":0.43,"tiers":[{"input":0.07,"output":0.43,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.07,"output":0.43}},"canonical_model_id":"google/gemini-3-flash-preview"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":0.71,"output":3.57},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5-Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0.04,"output":0.29},"canonical_model_id":"openai/gpt-5-mini"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.09,"output":0.71,"tiers":[{"input":0.09,"output":0.71,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":0.09,"output":0.71}},"canonical_model_id":"google/gemini-2.5-flash"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.25,"output":2},"canonical_model_id":"openai/gpt-5.2"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":0.43,"output":2.14},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":0.14,"output":0.71},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":0.14,"output":1.14},"canonical_model_id":"openai/gpt-5.2-codex"},"gemini-3-pro-preview":{"id":"gemini-3-pro-preview","name":"Gemini 3 Pro Preview","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-11","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65000},"cost":{"input":0.57,"output":3.43},"canonical_model_id":"google/gemini-3-pro-preview"}}},"abacus":{"id":"abacus","env":["ABACUS_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://routellm.abacus.ai/v1","name":"Abacus","doc":"https://abacus.ai/help/api","models":{"gpt-4.1-mini":{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"gemini-3.1-flash-image":{"id":"gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image"},"gpt-4.1-nano":{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"gpt-5.1":{"id":"gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"grok-4-0709":{"id":"grok-4-0709","name":"Grok 4","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16384},"status":"deprecated","cost":{"input":3,"output":15}},"gemini-3-flash-preview":{"id":"gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"gemini-3.5-flash":{"id":"gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15},"canonical_model_id":"google/gemini-3.5-flash"},"gpt-5.1-codex-max":{"id":"gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-codex-max"},"gpt-5.4":{"id":"gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"claude-opus-4-5-20251101":{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-5-20251101"},"gpt-5.3-codex-xhigh":{"id":"gpt-5.3-codex-xhigh","name":"GPT-5.3 Codex XHigh","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-codex"},"gemini-3.1-pro-preview":{"id":"gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3.1-pro-preview"},"gpt-4.1":{"id":"gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"gpt-5.1-codex":{"id":"gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"gemini-3.7-flash":{"id":"gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.7-flash"},"claude-sonnet-5":{"id":"claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2,"output":10},"canonical_model_id":"anthropic/claude-sonnet-5"},"claude-opus-4-1-20250805":{"id":"claude-opus-4-1-20250805","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75},"canonical_model_id":"anthropic/claude-opus-4-1-20250805"},"gpt-5-mini":{"id":"gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"claude-3-7-sonnet-20250219":{"id":"claude-3-7-sonnet-20250219","name":"Claude Sonnet 3.7","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-3-7-sonnet-20250219"},"grok-4-fast-non-reasoning":{"id":"grok-4-fast-non-reasoning","name":"Grok 4 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":0.5},"canonical_model_id":"xai/grok-4-fast-non-reasoning"},"kimi-k2.5":{"id":"kimi-k2.5","name":"Kimi K2.5","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"status":"deprecated","cost":{"input":0.6,"output":3},"canonical_model_id":"moonshotai/kimi-k2.5"},"grok-4.6":{"id":"grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":32768},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"gpt-5.5":{"id":"gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"gemini-2.5-flash-image":{"id":"gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768},"cost":{"input":0.3,"output":30},"canonical_model_id":"google/gemini-2.5-flash-image"},"claude-opus-5":{"id":"claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-5"},"claude-sonnet-4-6":{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"muse-spark-1.1":{"id":"muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32000},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"claude-fable-5":{"id":"claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50},"canonical_model_id":"anthropic/claude-fable-5"},"qwen-2.5-coder-32b":{"id":"qwen-2.5-coder-32b","name":"Qwen 2.5 Coder 32B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.79,"output":0.79},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"gpt-5.2-chat-latest":{"id":"gpt-5.2-chat-latest","name":"GPT-5.2 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.2-chat-latest"},"gpt-4o-2024-11-20":{"id":"gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"status":"deprecated","cost":{"input":2.5,"output":10},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"mimo-v2-pro":{"id":"mimo-v2-pro","name":"MiMo-V2-Pro","description":"Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1,"output":3,"cache_read":0.2},"canonical_model_id":"xiaomi/mimo-v2-pro"},"gpt-5-nano":{"id":"gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"claude-opus-4-6":{"id":"claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-6"},"grok-4.3":{"id":"grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"gemini-3-pro-image":{"id":"gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image"},"gpt-5.4-mini":{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"llama-3.3-70b-versatile":{"id":"llama-3.3-70b-versatile","name":"Llama 3.3 70B Versatile","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"status":"deprecated","cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"gemini-3.1-flash-lite-preview":{"id":"gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"status":"deprecated","cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":1},"canonical_model_id":"google/gemini-3.1-flash-lite-preview"},"grok-code-fast-1":{"id":"grok-code-fast-1","name":"Grok Code Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":1.5},"canonical_model_id":"xai/grok-code-fast-1"},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"google/gemini-2.5-pro"},"gpt-5.3-codex":{"id":"gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"o3-mini":{"id":"o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"gpt-5.2":{"id":"gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"gpt-5.6-terra":{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"openai/gpt-5.6-terra"},"claude-sonnet-4-20250514":{"id":"claude-sonnet-4-20250514","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"status":"deprecated","cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-20250514"},"claude-sonnet-4-5-20250929":{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15},"canonical_model_id":"anthropic/claude-sonnet-4-5-20250929"},"claude-opus-4-8":{"id":"claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-8"},"qwen3-max":{"id":"qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":16384},"status":"deprecated","cost":{"input":1.2,"output":6},"canonical_model_id":"alibaba/qwen3-max"},"route-llm":{"id":"route-llm","name":"RouteLLM","description":"RouteLLM routes prompts to an appropriate Abacus-backed text-generation model","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-01-01","last_updated":"2026-07-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000},"cost":{"input":2,"output":10,"cache_read":0.1}},"o4-mini":{"id":"o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4},"canonical_model_id":"openai/o4-mini"},"qwen3.7-max":{"id":"qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":2.5,"output":7.5},"canonical_model_id":"alibaba/qwen3.7-max"},"gpt-5.1-chat-latest":{"id":"gpt-5.1-chat-latest","name":"GPT-5.1 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10},"canonical_model_id":"openai/gpt-5.1-chat-latest"},"gpt-5.4-nano":{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"gemini-3.6-flash":{"id":"gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.6-flash"},"gpt-4o":{"id":"gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"grok-4-1-fast-non-reasoning":{"id":"grok-4-1-fast-non-reasoning","name":"Grok 4.1 Fast (Non-Reasoning)","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-11-17","last_updated":"2025-11-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":16384},"status":"deprecated","cost":{"input":0.2,"output":0.5}},"claude-haiku-4-5-20251001":{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5},"canonical_model_id":"anthropic/claude-haiku-4-5-20251001"},"o3":{"id":"o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768},"status":"deprecated","cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"gpt-5.6-luna":{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"qwen3.8-max":{"id":"qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.8-max"},"claude-opus-4-7":{"id":"claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25},"canonical_model_id":"anthropic/claude-opus-4-7"},"gpt-4o-mini":{"id":"gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-4o-mini"},"claude-opus-4-20250514":{"id":"claude-opus-4-20250514","name":"Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"status":"deprecated","cost":{"input":15,"output":75},"canonical_model_id":"anthropic/claude-opus-4-20250514"},"gpt-5.2-codex":{"id":"gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"gemini-3.5-flash-lite":{"id":"gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-3.5-flash-lite"},"gpt-5":{"id":"gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"o3-pro":{"id":"o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":20,"output":80},"canonical_model_id":"openai/o3-pro"},"muse-spark-1.2":{"id":"muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"gpt-5-codex":{"id":"gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"status":"deprecated","cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5-codex"},"gpt-5.3-chat-latest":{"id":"gpt-5.3-chat-latest","name":"GPT-5.3 Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"output":128000},"status":"deprecated","cost":{"input":1.75,"output":14},"canonical_model_id":"openai/gpt-5.3-chat-latest"},"grok-4.5":{"id":"grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":32768},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"gemini-3.1-flash-lite":{"id":"gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025},"canonical_model_id":"google/gemini-3.1-flash-lite"},"gpt-5.6-sol":{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4},"canonical_model_id":"openai/gpt-5.6-sol"},"gemini-3.1-flash-image-preview":{"id":"gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":32768},"status":"deprecated","cost":{"input":0.5,"output":3},"canonical_model_id":"google/gemini-3.1-flash-image-preview"},"kimi-k2-turbo-preview":{"id":"kimi-k2-turbo-preview","name":"Kimi K2 Turbo Preview","description":"Fast Kimi model for responsive chat, coding help, and agent loops","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-08","last_updated":"2025-07-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":8192},"status":"deprecated","cost":{"input":0.15,"output":8}},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":0.14,"output":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"MiniMaxAI/MiniMax-M3":{"id":"MiniMaxAI/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M3"},"MiniMaxAI/MiniMax-M2.7":{"id":"MiniMaxAI/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.3,"output":1.2},"canonical_model_id":"minimax/MiniMax-M2.7"},"thinkingmachines/Inkling":{"id":"thinkingmachines/Inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":3.74,"output":9.36,"cache_read":0.748},"canonical_model_id":"thinkingmachines/inkling"},"meta-llama/Meta-Llama-3.1-8B-Instruct":{"id":"meta-llama/Meta-Llama-3.1-8B-Instruct","name":"Llama 3.1 8B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":0.02,"output":0.05},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/Meta-Llama-3.3-70B-Instruct":{"id":"meta-llama/Meta-Llama-3.3-70B-Instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"status":"deprecated","cost":{"input":0.59,"output":0.79},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo":{"id":"meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo","name":"Llama 3.1 405B Instruct Turbo","description":"Compact Llama instruction model for fast chat and local deployment","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"status":"deprecated","cost":{"input":3.5,"output":3.5}},"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8":{"id":"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192},"status":"deprecated","cost":{"input":0.14,"output":0.59},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.55,"output":1.66},"canonical_model_id":"deepseek/deepseek-v3.1"},"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.16},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/Kimi-K3":{"id":"moonshotai/Kimi-K3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"moonshotai/kimi-k3"},"Qwen/Qwen3-Coder-480B-A35B-Instruct":{"id":"Qwen/Qwen3-Coder-480B-A35B-Instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"cost":{"input":0.38,"output":1.55},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"Qwen/Qwen3-235B-A22B-Instruct-2507":{"id":"Qwen/Qwen3-235B-A22B-Instruct-2507","name":"Qwen3 235B A22B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"status":"deprecated","cost":{"input":0.13,"output":0.6},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"Qwen/Qwen3-32B":{"id":"Qwen/Qwen3-32B","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"cost":{"input":0.08,"output":0.28},"canonical_model_id":"alibaba/qwen3-32b"},"Qwen/Qwen3.6-27B":{"id":"Qwen/Qwen3.6-27B","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.32,"output":3.2},"canonical_model_id":"alibaba/qwen3.6-27b"},"Qwen/Qwen2.5-72B-Instruct":{"id":"Qwen/Qwen2.5-72B-Instruct","name":"Qwen 2.5 72B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.11,"output":0.38}},"Qwen/QwQ-32B":{"id":"Qwen/QwQ-32B","name":"QwQ 32B","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"status":"deprecated","cost":{"input":0.4,"output":0.4},"canonical_model_id":"alibaba/qwq-32b"},"deepseek-ai/DeepSeek-R1":{"id":"deepseek-ai/DeepSeek-R1","name":"DeepSeek R1","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":3,"output":7},"canonical_model_id":"deepseek/deepseek-r1"},"deepseek-ai/DeepSeek-V4-Pro":{"id":"deepseek-ai/DeepSeek-V4-Pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"status":"deprecated","cost":{"input":1.74,"output":3.48,"cache_read":0.15},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek-ai/DeepSeek-V4-Flash":{"id":"deepseek-ai/DeepSeek-V4-Flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":32768},"status":"deprecated","cost":{"input":0.14,"output":0.28,"cache_read":0.03},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2025-06-01","last_updated":"2025-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.27,"output":1}},"deepseek-ai/DeepSeek-V3.2":{"id":"deepseek-ai/DeepSeek-V3.2","name":"DeepSeek V3.2","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192},"status":"deprecated","cost":{"input":0.27,"output":0.4},"canonical_model_id":"deepseek/deepseek-v3.2"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"cost":{"input":0.037,"output":0.17},"canonical_model_id":"openai/gpt-oss-120b"},"zai-org/GLM-4.5":{"id":"zai-org/GLM-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":96000},"status":"deprecated","cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.5"},"zai-org/GLM-4.6":{"id":"zai-org/GLM-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"output":131072},"cost":{"input":0.55,"output":2.2},"canonical_model_id":"zhipuai/glm-4.6"},"zai-org/GLM-5":{"id":"zai-org/GLM-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1,"output":3.2},"canonical_model_id":"zhipuai/glm-5"},"zai-org/GLM-5.1":{"id":"zai-org/GLM-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":1.38,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.1"},"zai-org/GLM-4.7":{"id":"zai-org/GLM-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"cost":{"input":0.6,"output":2.2},"canonical_model_id":"zhipuai/glm-4.7"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.2"}}},"claudinio":{"id":"claudinio","env":["CLAUDINIO_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.claudin.io/v1","name":"Claudinio","doc":"https://claudin.io","models":{"claudius":{"id":"claudius","name":"Claudius","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"knowledge":"2026-05","release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":3,"output":8,"cache_read":0.9}},"claudinio":{"id":"claudinio","name":"Claudinio","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"knowledge":"2026-05","release_date":"2026-05-12","last_updated":"2026-06-02","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000},"cost":{"input":0.5,"output":2,"cache_read":0.15}}}},"nearai":{"id":"nearai","env":["NEARAI_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://cloud-api.near.ai/v1","name":"NEAR AI Cloud","doc":"https://docs.near.ai/","models":{"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":1.8,"output":5.75,"cache_read":0.45},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"cost":{"input":0.08,"output":0.5,"cache_read":0.012},"canonical_model_id":"zhipuai/glm-5.3-flash"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"input_audio":1.5},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"input_audio":1},"canonical_model_id":"google/gemini-2.5-flash"},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"tiers":[{"input":2.5,"output":15,"cache_read":0.25,"tier":{"type":"context","size":200000}}],"context_over_200k":{"input":2.5,"output":15,"cache_read":0.25}},"canonical_model_id":"google/gemini-2.5-pro"},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"input_audio":0.5},"canonical_model_id":"google/gemini-3.1-flash-lite"},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01,"input_audio":0.3},"canonical_model_id":"google/gemini-2.5-flash-lite"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"cost":{"input":0.3,"output":1.2,"cache_read":0.006},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"cost":{"input":1.1,"output":1.1,"cache_read":0.55},"canonical_model_id":"deepseek/deepseek-v3.2"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":3.3,"output":16.5,"cache_read":0.33},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.81,"output":3.85,"cache_read":0.41},"canonical_model_id":"moonshotai/kimi-k2.6"},"Qwen/Qwen3-VL-30B-A3B-Instruct":{"id":"Qwen/Qwen3-VL-30B-A3B-Instruct","name":"Qwen3-VL 30B-A3B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":8192},"cost":{"input":0.15,"output":0.55,"cache_read":0.03},"canonical_model_id":"alibaba/qwen3-vl-30b-a3b-instruct"},"Qwen/Qwen3.6-35B-A3B-FP8":{"id":"Qwen/Qwen3.6-35B-A3B-FP8","name":"Qwen 3.6 35B A3B FP8","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":8192},"cost":{"input":0.17,"output":1.1,"cache_read":0.056},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"Qwen/Qwen3-Reranker-0.6B":{"id":"Qwen/Qwen3-Reranker-0.6B","type":"reranking","name":"Qwen3 Reranker 0.6B","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":40960,"output":1024},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"alibaba/qwen3-reranker-0.6b"},"Qwen/Qwen3.8-27B":{"id":"Qwen/Qwen3.8-27B","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.04,"output":1.35,"cache_read":0.018},"canonical_model_id":"alibaba/qwen3.8-27b"},"Qwen/Qwen3-Embedding-0.6B":{"id":"Qwen/Qwen3-Embedding-0.6B","type":"embedding","name":"Qwen3 Embedding 0.6B","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":1024},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"alibaba/qwen3-embedding-0.6b"},"black-forest-labs/FLUX.2-klein-4B":{"id":"black-forest-labs/FLUX.2-klein-4B","type":"image","name":"FLUX.2 Klein 4B","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-15","last_updated":"2026-01-15","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"canonical_model_id":"black-forest-labs/flux-2-klein-4b"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25,"tiers":[{"input":5,"output":22.5,"cache_read":0.5,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":5,"output":22.5,"cache_read":0.5}},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5,"tiers":[{"input":10,"output":45,"cache_read":1,"tier":{"type":"context","size":272000}}],"context_over_200k":{"input":10,"output":45,"cache_read":1}},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.275},"canonical_model_id":"openai/o4-mini"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/o3"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/whisper-large-v3":{"id":"openai/whisper-large-v3","type":"transcription","name":"Whisper Large v3","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448},"cost":{"input":0.01,"output":0.01},"canonical_model_id":"openai/whisper-large-v3"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.4},"canonical_model_id":"openai/gpt-5.6-sol"},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3,"cache_write":3.75},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1,"cache_write":1.25},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"budget_tokens","min":1024}],"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-4-7"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"cost":{"input":0.5,"output":3.3,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"cost":{"input":2.8,"output":7.5},"canonical_model_id":"alibaba/qwen3.7-max"},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"}}},"nano-gpt":{"id":"nano-gpt","env":["NANO_GPT_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://nano-gpt.com/api/v1","name":"NanoGPT","doc":"https://docs.nano-gpt.com","models":{"gemma-4-31b-it-novelist":{"id":"gemma-4-31b-it-novelist","name":"Novelist","description":"Novelist is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"doubao-seed-2-0-lite-260215":{"id":"doubao-seed-2-0-lite-260215","name":"Doubao Seed 2.0 Lite","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32000},"cost":{"input":0.1462,"output":0.8738,"cache_read":0.0731},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"gemma-4-31b-it-scarlet":{"id":"gemma-4-31b-it-scarlet","name":"Gemma 4 31B Scarlet","description":"Scarlet is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-08","last_updated":"2026-10-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"ernie-5.0-thinking-preview":{"id":"ernie-5.0-thinking-preview","name":"Ernie 5.0 Thinking Preview","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"ernie","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":1,"output":3.5,"cache_read":0.5}},"ernie-5.1:thinking":{"id":"ernie-5.1:thinking","name":"ERNIE 5.1 Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"ernie","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2026-05-10","last_updated":"2026-05-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":119000,"input":119000,"output":64000},"cost":{"input":0.75,"output":3,"cache_read":0.75}},"mercury-2":{"id":"mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":50000},"cost":{"input":0.25,"output":0.75,"cache_read":0.025},"canonical_model_id":"inception/mercury-2"},"gemini-2.5-flash-lite-preview-06-17":{"id":"gemini-2.5-flash-lite-preview-06-17","name":"Gemini 2.5 Flash Lite Preview","description":"First preview of Gemini 2.5 Flash-Lite, the low-cost, low-latency 2.5 model","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-06-17"},"qvq-max":{"id":"qvq-max","name":"Qwen: QvQ Max","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-03-28","last_updated":"2025-03-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":1.2,"output":4.8,"cache_read":0.6}},"phi-4-mini-instruct":{"id":"phi-4-mini-instruct","name":"Phi 4 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.17,"output":0.68,"cache_read":0.085},"canonical_model_id":"microsoft/phi-4-mini"},"celeris-1":{"id":"celeris-1","name":"Celeris 1","description":"Celeris 1 is a diffusion language model built for ultra-low-latency classification, extraction, judging, query rewriting, and other short structured responses.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-07-25","last_updated":"2026-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":2,"output":6,"cache_read":1}},"gemma-4-31b-it-garnet":{"id":"gemma-4-31b-it-garnet","name":"Garnet","description":"Garnet is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"hermes-low":{"id":"hermes-low","name":"Hermes Low","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2,"cache_write":0.08333}},"nano-gpt-help":{"id":"nano-gpt-help","name":"NanoGPT Help","description":"Text-only NanoGPT support assistant. Questions are processed by the Help inference provider; do not paste secrets or account credentials. Covers the website, models, API, pricing, memory, media generation, and support.","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-06-06","last_updated":"2026-06-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":6000,"input":6000,"output":512},"cost":{"input":0,"output":0}},"longcat-2.5-preview":{"id":"longcat-2.5-preview","name":"LongCat 2.5 Preview","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.75,"output":2.95,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.5-preview"},"gemini-2.5-flash-preview-09-2025":{"id":"gemini-2.5-flash-preview-09-2025","name":"Gemini 2.5 Flash Preview (09/2025)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"Gemma-4-31B-Cognitive-Unshackled":{"id":"Gemma-4-31B-Cognitive-Unshackled","name":"Gemma 4 31B Cognitive Unshackled","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"gemma-4-26b-a4b-it-shadowsiren":{"id":"gemma-4-26b-a4b-it-shadowsiren","name":"Shadow Siren","description":"Shadow Siren is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"deepseek-v3-0324":{"id":"deepseek-v3-0324","name":"DeepSeek Chat 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.2,"output":0.77,"cache_read":0.135},"canonical_model_id":"deepseek/deepseek-v3-0324"},"GLM-4.6-Derestricted-v5":{"id":"GLM-4.6-Derestricted-v5","name":"GLM 4.6 Derestricted v5","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":8192},"cost":{"input":0.4,"output":1.5,"cache_read":0.2}},"doubao-seed-2-0-mini-260215":{"id":"doubao-seed-2-0-mini-260215","name":"Doubao Seed 2.0 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32000},"cost":{"input":0.0493,"output":0.4845,"cache_read":0.02465},"canonical_model_id":"bytedance-seed/seed-2.0-mini"},"glm-4-long":{"id":"glm-4-long","name":"GLM-4 Long","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":4095},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003}},"doubao-seed-2-0-pro-260215":{"id":"doubao-seed-2-0-pro-260215","name":"Doubao Seed 2.0 Pro","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":0.782,"output":3.876,"cache_read":0.391},"canonical_model_id":"bytedance-seed/seed-2.0-pro"},"auto-model":{"id":"auto-model","name":"Auto model","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":0,"output":0}},"pokee-isaac":{"id":"pokee-isaac","name":"Pokee-Isaac 28B","description":"Pokee-Isaac is a 28B agentic model with a roughly 10-million-token context window, function calling, and OpenAI-compatible structured output. Pokee bills in $0.01 increments, rounding each non-zero request up to the next cent.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-08-04","last_updated":"2026-08-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"input":10000000,"output":60000},"cost":{"input":0.15,"output":1,"cache_read":0.075}},"fastgpt":{"id":"fastgpt","name":"Web Answer","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-23","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":7.5,"output":7.5}},"Gemma-4-31B-Claude-4.6-Opus-Reasoning-Distilled":{"id":"Gemma-4-31B-Claude-4.6-Opus-Reasoning-Distilled","name":"Gemma 4 31B Claude 4.6 Opus Reasoning Distilled","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"claude","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.0306}},"gemma-4-31b-it-gemsicle":{"id":"gemma-4-31b-it-gemsicle","name":"Gemsicle","description":"Gemsicle is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"Gemma-4-26B-A4B-MeroMero":{"id":"Gemma-4-26B-A4B-MeroMero","name":"Gemma 4 26B A4B MeroMero","description":"Gemma 4 26B A4B MeroMero is an NVFP4 multimodal mixture-of-experts fine-tune for emotive dialogue, relationship scenes, creative writing, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"doubao-1.5-pro-32k":{"id":"doubao-1.5-pro-32k","name":"Doubao 1.5 Pro 32k","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-11-20","last_updated":"2025-01-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.1343,"output":0.3349,"cache_read":0.06715}},"qwen3-vl-235b-a22b-instruct-original":{"id":"qwen3-vl-235b-a22b-instruct-original","name":"Qwen3 VL 235B A22B Instruct Original","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.3,"output":1.9,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"holo3-35b-a3b":{"id":"holo3-35b-a3b","name":"Holo3-35B-A3B","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.25,"output":1.8,"cache_read":0.125}},"venice-uncensored":{"id":"venice-uncensored","name":"Venice Uncensored","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"venice","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-10-01","last_updated":"2025-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.4,"output":1.8,"cache_read":0.4}},"hermes-high":{"id":"hermes-high","name":"Hermes High","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"auto-model-premium":{"id":"auto-model-premium","name":"Auto model (Premium)","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":9.996,"output":19.992,"cache_read":4.998}},"gemini-2.5-flash":{"id":"gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03},"canonical_model_id":"google/gemini-2.5-flash"},"gemma-4-26b-a4b-it-musica":{"id":"gemma-4-26b-a4b-it-musica","name":"Musica","description":"Musica is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemma-4-12b-it-semancer":{"id":"gemma-4-12b-it-semancer","name":"Gemma 4 12B Semancer","description":"Gemma 4 12B Semancer is an open-weight roleplay finetune with image understanding, tool calling, optional reasoning, and a 131,072-token context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.025}},"glm-4.1v-thinking-flash":{"id":"glm-4.1v-thinking-flash","name":"GLM 4.1V Thinking Flash","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":64000,"input":64000,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"Gemma-4-31B-GarnetV2":{"id":"Gemma-4-31B-GarnetV2","name":"Gemma 4 31B Garnet V2","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"longcat-2.0:thinking":{"id":"longcat-2.0:thinking","name":"LongCat 2.0 Thinking","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.75,"output":3,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"mistral-code-latest":{"id":"mistral-code-latest","name":"Mistral Code Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.15}},"doubao-1.5-pro-256k":{"id":"doubao-1.5-pro-256k","name":"Doubao 1.5 Pro 256k","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":16384},"cost":{"input":0.799,"output":1.445,"cache_read":0.3995}},"gemini-2.5-pro-preview-05-06":{"id":"gemini-2.5-pro-preview-05-06","name":"Gemini 2.5 Pro Preview 0506","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-05-06","last_updated":"2025-05-06","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25}},"phi-4-multimodal-instruct":{"id":"phi-4-multimodal-instruct","name":"Phi 4 Multimodal","description":"Compact Microsoft multimodal model for text, image, and speech understanding","family":"phi","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.07,"output":0.11,"cache_read":0.035},"canonical_model_id":"microsoft/phi-4-multimodal-instruct"},"gemma-4-26b-a4b-it-luminous":{"id":"gemma-4-26b-a4b-it-luminous","name":"Luminous Mirror","description":"Luminous Mirror is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"kimi-k2-instruct-fast":{"id":"kimi-k2-instruct-fast","name":"Kimi K2 0711 Fast","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"status":"deprecated","cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"Gemma-4-31B-MeroMero-v2":{"id":"Gemma-4-31B-MeroMero-v2","name":"Gemma 4 31B MeroMero v2","description":"Gemma 4 31B MeroMero v2 is a LoRA finetune for emotive dialogue, relationship scenes, creative writing, and multimodal roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemini-2.5-flash-preview-09-2025-thinking":{"id":"gemini-2.5-flash-preview-09-2025-thinking","name":"Gemini 2.5 Flash Preview (09/2025) – Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"gemma-4-26b-a4b-it-chimerax":{"id":"gemma-4-26b-a4b-it-chimerax","name":"Chimera X","description":"Chimera X is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemini-2.5-pro":{"id":"gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.25,"output":10,"cache_read":0.125,"cache_write":0.375},"canonical_model_id":"google/gemini-2.5-pro"},"Gemma-4-31B-Queen":{"id":"Gemma-4-31B-Queen","name":"Gemma 4 31B Queen","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"longcat-2.0":{"id":"longcat-2.0","name":"LongCat 2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.75,"output":3,"cache_read":0.015},"canonical_model_id":"meituan/longcat-2.0"},"command-a-reasoning-08-2025":{"id":"command-a-reasoning-08-2025","name":"Cohere Command A (08/2025)","description":"Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows","family":"command-a","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":8192},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"cohere/command-a-reasoning-08-2025"},"gemini-2.0-pro-exp-02-05":{"id":"gemini-2.0-pro-exp-02-05","name":"Gemini 2.0 Pro 0205","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-05","last_updated":"2025-02-05","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":2097152,"input":2097152,"output":8192},"cost":{"input":1.989,"output":7.956,"cache_read":0.49725}},"deepseek-chat-cheaper":{"id":"deepseek-chat-cheaper","name":"DeepSeek V3/Chat Cheaper","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.1,"output":0.425,"cache_read":0.05}},"Gemma-4-31B-DarkIdol":{"id":"Gemma-4-31B-DarkIdol","name":"Gemma 4 31B DarkIdol","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-05-01","last_updated":"2026-05-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"glm-4-plus-0111":{"id":"glm-4-plus-0111","name":"GLM 4 Plus 0111","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":9.996,"output":9.996,"cache_read":4.998}},"ernie-5.1":{"id":"ernie-5.1","name":"ERNIE 5.1","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"ernie","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-05-10","last_updated":"2026-05-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":119000,"input":119000,"output":64000},"cost":{"input":0.75,"output":3,"cache_read":0.75}},"holo3-35b-a3b:thinking":{"id":"holo3-35b-a3b:thinking","name":"Holo3-35B-A3B Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.25,"output":1.8,"cache_read":0.125}},"glm-4-air-0111":{"id":"glm-4-air-0111","name":"GLM 4 Air 0111","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-01-11","last_updated":"2025-01-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":0.1394,"output":0.1394,"cache_read":0.0697}},"asi1-mini":{"id":"asi1-mini","name":"ASI1 Mini","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":1,"output":1,"cache_read":0.5}},"mercury-coder-small":{"id":"mercury-coder-small","name":"Mercury Coder Small","description":"Model by Inception AI. A diffusion large language model that runs incredibly quickly (500+ tokens/second) while matching Claude 3.5 Haiku and GPT-4o-mini. 1st in speed on Copilot arena, and matching 2nd in quality.","family":"mercury","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.25,"output":1,"cache_read":0.125}},"gemma-4-31b-it-darkidol":{"id":"gemma-4-31b-it-darkidol","name":"DarkIdol","description":"DarkIdol is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemini-2.0-pro-reasoner":{"id":"gemini-2.0-pro-reasoner","name":"Gemini 2.0 Pro Reasoner","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-02-05","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":1.292,"output":4.998,"cache_read":0.323}},"gemma-4-12b-it":{"id":"gemma-4-12b-it","name":"Gemma 4 12B Instruct","description":"Compact Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.025},"canonical_model_id":"google/gemma-4-12b-it"},"gemma-4-26b-a4b-it-darksoul":{"id":"gemma-4-26b-a4b-it-darksoul","name":"Dark Soul","description":"Dark Soul is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemini-exp-1206":{"id":"gemini-exp-1206","name":"Gemini 2.0 Pro 1206","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":2097152,"input":2097152,"output":8192},"cost":{"input":1.258,"output":4.998,"cache_read":0.629}},"gemini-2.5-pro-preview-03-25":{"id":"gemini-2.5-pro-preview-03-25","name":"Gemini 2.5 Pro Preview 0325","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25}},"gemma-4-31b-it-isometry":{"id":"gemma-4-31b-it-isometry","name":"Isometry","description":"Isometry is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"gemma-4-31b-it-gembrain":{"id":"gemma-4-31b-it-gembrain","name":"Gembrain","description":"Gembrain is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"glm-z1-airx":{"id":"glm-z1-airx","name":"GLM Z1 AirX","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":16384},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"doubao-1.5-vision-pro-32k":{"id":"doubao-1.5-vision-pro-32k","name":"Doubao 1.5 Vision Pro 32k","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-11-20","last_updated":"2025-01-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.459,"output":1.377,"cache_read":0.2295}},"auto-model-standard":{"id":"auto-model-standard","name":"Auto model (Standard)","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":9.996,"output":19.992,"cache_read":4.998}},"Gemma-4-31B-MeroMero-v2:thinking":{"id":"Gemma-4-31B-MeroMero-v2:thinking","name":"Gemma 4 31B MeroMero v2 Thinking","description":"Gemma 4 31B MeroMero v2 with thinking enabled for more deliberate emotive dialogue, relationship scenes, creative writing, and multimodal roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"sarvam-105b":{"id":"sarvam-105b","name":"Sarvam 105B","description":"Flagship Indian-language reasoning model for enterprise multilingual applications","family":"sarvam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":4096},"cost":{"input":0.054,"output":0.2124,"cache_read":0.0336},"canonical_model_id":"sarvam/sarvam-105b"},"MiniMax-M1":{"id":"MiniMax-M1","name":"MiniMax M1","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-01-08","last_updated":"2025-06-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.1394,"output":1.3328,"cache_read":0.0697}},"claw-medium":{"id":"claw-medium","name":"Claw Medium","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"gemma-4-12b-it-station-keeper":{"id":"gemma-4-12b-it-station-keeper","name":"Gemma 4 12B StationKeeper","description":"Gemma 4 12B StationKeeper is an open-weight roleplay finetune with image understanding, tool calling, optional reasoning, and a 131,072-token context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.25,"cache_read":0.025}},"glm-4.1v-thinking-flashx":{"id":"glm-4.1v-thinking-flashx","name":"GLM 4.1V Thinking FlashX","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"glm-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":64000,"input":64000,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"mistral-code-agent-latest":{"id":"mistral-code-agent-latest","name":"Mistral Code Agent Latest","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.4,"output":2,"cache_read":0.2}},"hermes-medium":{"id":"hermes-medium","name":"Hermes Medium","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"hermes","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"deepseek-chat":{"id":"deepseek-chat","name":"DeepSeek V3/Deepseek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.1,"output":0.425,"cache_read":0.05},"canonical_model_id":"deepseek/deepseek-chat"},"Qwen3.5-27B-Queen-Derestricted":{"id":"Qwen3.5-27B-Queen-Derestricted","name":"Qwen3.5 27B Queen Derestricted","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"gemini-2.5-pro-exp-03-25":{"id":"gemini-2.5-pro-exp-03-25","name":"Gemini 2.5 Pro Experimental 0325","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-03-25","last_updated":"2025-03-25","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25}},"Meta-Llama-3-1-8B-Instruct-FP8":{"id":"Meta-Llama-3-1-8B-Instruct-FP8","name":"Llama 3.1 8B (decentralized)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.02,"output":0.03,"cache_read":0.01}},"Gemma-4-26B-A4B-MeroMero:thinking":{"id":"Gemma-4-26B-A4B-MeroMero:thinking","name":"Gemma 4 26B A4B MeroMero Thinking","description":"Gemma 4 26B A4B MeroMero with thinking enabled for more deliberate emotive dialogue, relationship scenes, creative writing, and multimodal roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"doubao-seed-2-0-code-preview-260215":{"id":"doubao-seed-2-0-code-preview-260215","name":"Doubao Seed 2.0 Code Preview","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":0.782,"output":3.893,"cache_read":0.391},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"gemini-2.5-flash-lite-preview-09-2025":{"id":"gemini-2.5-flash-lite-preview-09-2025","name":"Gemini 2.5 Flash Lite Preview (09/2025)","description":"September 2025 preview of Gemini 2.5 Flash-Lite with better instruction following and multimodal quality","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-09-2025"},"gemini-2.5-flash-preview-05-20":{"id":"gemini-2.5-flash-preview-05-20","name":"Gemini 2.5 Flash 0520","description":"Preview snapshot of Gemini 2.5 Flash with controllable thinking budgets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"input":1048000,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"google/gemini-2.5-flash-preview-05-20"},"gemini-2.5-flash-lite-preview-09-2025-thinking":{"id":"gemini-2.5-flash-lite-preview-09-2025-thinking","name":"Gemini 2.5 Flash Lite Preview (09/2025) – Thinking","description":"September 2025 preview of Gemini 2.5 Flash-Lite with better instruction following and multimodal quality","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite-preview-09-2025"},"gemma-4-26b-a4b-it-chimeraxheretic":{"id":"gemma-4-26b-a4b-it-chimeraxheretic","name":"Gemma 4 26B A4B Chimera X Heretic","description":"Chimera X Heretic is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-08","last_updated":"2026-10-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemini-3-pro-image-preview":{"id":"gemini-3-pro-image-preview","type":"image","name":"Gemini 3 Pro Image","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"input":65536,"output":32768},"cost":{"input":2,"output":12,"cache_read":0.2},"canonical_model_id":"google/gemini-3-pro-image-preview"},"gemma-4-26b-a4b-it-moonlight":{"id":"gemma-4-26b-a4b-it-moonlight","name":"Moonlight Dusk","description":"Moonlight Dusk is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemma-4-31b-it-fabled":{"id":"gemma-4-31b-it-fabled","name":"Fabled","description":"Fabled is a multimodal Gemma 4 31B creative finetune for expressive dialogue, long-form storytelling, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"universal-summarizer":{"id":"universal-summarizer","name":"Universal Summarizer","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-23","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":30,"output":30}},"agnes-3.0-flash":{"id":"agnes-3.0-flash","name":"Agnes 3.0 Flash","description":"Agnes 3.0 Flash is a low-cost model for coding, tool use, and multi-turn agent tasks. It supports text and image input, optional thinking, and a 512K-token context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":65536},"cost":{"input":0.05,"output":0.15,"cache_read":0.005}},"gemini-2.5-flash-preview-04-17:thinking":{"id":"gemini-2.5-flash-preview-04-17:thinking","name":"Gemini 2.5 Flash Preview Thinking","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-04-17","last_updated":"2025-04-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":3.5,"cache_read":0.015}},"Qwen3.5-27B-BlueStar-v3-Derestricted":{"id":"Qwen3.5-27B-BlueStar-v3-Derestricted","name":"Qwen3.5 27B BlueStar v3 Derestricted","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"qwen3.5","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.306,"output":0.306,"cache_read":0.153}},"claw-high":{"id":"claw-high","name":"Claw High","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"auto-model-basic":{"id":"auto-model-basic","name":"Auto model (Basic)","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-04-16","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":9.996,"output":19.992,"cache_read":4.998}},"gemini-2.5-pro-preview-06-05":{"id":"gemini-2.5-pro-preview-06-05","name":"Gemini 2.5 Pro Preview 0605","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2.5,"output":10,"cache_read":0.25},"canonical_model_id":"google/gemini-2.5-pro-preview"},"claw-low":{"id":"claw-low","name":"Claw Low","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-11","last_updated":"2026-05-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2,"cache_write":0.08333}},"gemini-2.5-flash-lite":{"id":"gemini-2.5-flash-lite","name":"Gemini 2.5 Flash Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.01},"canonical_model_id":"google/gemini-2.5-flash-lite"},"gemini-2.5-flash-nothinking":{"id":"gemini-2.5-flash-nothinking","name":"Gemini 2.5 Flash (No Thinking)","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03}},"gemini-2.5-flash-preview-05-20:thinking":{"id":"gemini-2.5-flash-preview-05-20:thinking","name":"Gemini 2.5 Flash 0520 Thinking","description":"Preview snapshot of Gemini 2.5 Flash with controllable thinking budgets","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048000,"input":1048000,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":3.5,"cache_read":0.015},"canonical_model_id":"google/gemini-2.5-flash-preview-05-20"},"gemma-4-26b-a4b-it-opusdistill":{"id":"gemma-4-26b-a4b-it-opusdistill","name":"Opus Distill","description":"Opus Distill is a Gemma 4 26B A4B multimodal mixture-of-experts fine-tune for creative writing, expressive dialogue, and roleplay.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-07-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.12,"output":0.38,"cache_read":0.06}},"gemini-2.5-flash-preview-04-17":{"id":"gemini-2.5-flash-preview-04-17","name":"Gemini 2.5 Flash Preview","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"release_date":"2025-04-17","last_updated":"2025-04-17","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"status":"deprecated","cost":{"input":0.15,"output":0.6,"cache_read":0.015}},"Gryphe/MythoMax-L2-13b":{"id":"Gryphe/MythoMax-L2-13b","name":"MythoMax 13B","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"input":4096,"output":3686},"cost":{"input":0.1003,"output":0.1003,"cache_read":0.05015},"canonical_model_id":"gryphe/mythomax-l2-13b"},"mistralai/devstral-small-2505":{"id":"mistralai/devstral-small-2505","name":"Mistral Devstral Small 2505","description":"OpenHands+Devstral is 100% local 100% open, and is SOTA for the category on SWE-Bench Verified: 46.8% accuracy.","family":"devstral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-08-02","last_updated":"2025-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.06,"output":0.06,"cache_read":0.03}},"mistralai/ministral-14b-2512":{"id":"mistralai/ministral-14b-2512","name":"Ministral 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.2,"output":0.2,"cache_read":0.1},"canonical_model_id":"mistral/ministral-14b-2512"},"mistralai/mixtral-8x22b-instruct-v0.1":{"id":"mistralai/mixtral-8x22b-instruct-v0.1","name":"Mixtral 8x22B","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":52428},"cost":{"input":2,"output":6,"cache_read":0.2},"canonical_model_id":"mistral/mixtral-8x22b-instruct"},"mistralai/mistral-small-4-119b-2603:thinking":{"id":"mistralai/mistral-small-4-119b-2603:thinking","name":"Mistral Small 4 119B Thinking","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.4,"output":1.4,"cache_read":0.2},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mistral-medium-3.1":{"id":"mistralai/mistral-medium-3.1","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.4,"output":2,"cache_read":0.2},"canonical_model_id":"mistral/mistral-medium-3.1"},"mistralai/codestral-2508":{"id":"mistralai/codestral-2508","name":"Codestral 2508","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.3,"output":0.9,"cache_read":0.15},"canonical_model_id":"mistral/codestral-2508"},"mistralai/mistral-small-3.2-24b-instruct":{"id":"mistralai/mistral-small-3.2-24b-instruct","name":"Mistral Small 3.2 24B (2506)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.2,"output":0.4,"cache_read":0.1},"canonical_model_id":"mistral/mistral-small-3.2-24b-instruct"},"mistralai/mistral-large":{"id":"mistralai/mistral-large","name":"Mistral Large 2411","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":102400},"cost":{"input":2.006,"output":6.001,"cache_read":0.2},"canonical_model_id":"mistral/mistral-large"},"mistralai/mistral-small-3.1-24b-instruct":{"id":"mistralai/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B (2503)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":102400},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"mistral/mistral-small-3.1-24b-instruct"},"mistralai/ministral-8b-2512":{"id":"mistralai/ministral-8b-2512","name":"Ministral 8B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.15,"output":0.15,"cache_read":0.075},"canonical_model_id":"mistral/ministral-8b-2512"},"mistralai/mistral-large-4":{"id":"mistralai/mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":262144},"cost":{"input":0.68,"output":2.09,"cache_read":0.07},"canonical_model_id":"mistral/mistral-large-4"},"mistralai/mistral-medium-3.5:thinking":{"id":"mistralai/mistral-medium-3.5:thinking","name":"Mistral Medium 3.5 Thinking","description":"Mistral Medium 3.5 with reasoning enabled by default (reasoning_effort=high), for complex coding, agentic, and multi-step reasoning prompts.","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.5,"output":7.5,"cache_read":0.75}},"mistralai/mistral-saba":{"id":"mistralai/mistral-saba","name":"Mistral Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":26214},"cost":{"input":0.1989,"output":0.595,"cache_read":0.09945},"canonical_model_id":"mistral/mistral-saba"},"mistralai/ministral-14b-instruct-2512":{"id":"mistralai/ministral-14b-instruct-2512","name":"Ministral 3 14B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"mistral/ministral-3-14b-instruct-2512"},"mistralai/mistral-nemo-instruct-2407":{"id":"mistralai/mistral-nemo-instruct-2407","name":"Mistral Nemo","description":"12B parameter model with multilingual support.","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.1003,"output":0.1207,"cache_read":0.05015},"canonical_model_id":"mistral/mistral-nemo"},"mistralai/devstral-2-123b-instruct-2512":{"id":"mistralai/devstral-2-123b-instruct-2512","name":"Devstral 2 123B","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.4,"output":1.4,"cache_read":0.2},"canonical_model_id":"mistral/devstral-2512"},"mistralai/mistral-small-24b-instruct-2501":{"id":"mistralai/mistral-small-24b-instruct-2501","name":"Mistral Small 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.1155,"output":0.3465},"canonical_model_id":"mistral/mistral-small-24b-instruct-2501"},"mistralai/mistral-small-4-119b-2603":{"id":"mistralai/mistral-small-4-119b-2603","name":"Mistral Small 4 119B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.4,"output":1.4,"cache_read":0.2},"canonical_model_id":"mistral/mistral-small-2603"},"mistralai/mistral-medium-3":{"id":"mistralai/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.4,"output":2,"cache_read":0.2},"canonical_model_id":"mistral/mistral-medium-3"},"mistralai/mistral-medium-3.5":{"id":"mistralai/mistral-medium-3.5","name":"Mistral Medium 3.5","description":"Mistral Medium 3.5 is a 128B dense open-weights flagship model for instruction-following, reasoning, coding, long-horizon agentic work, tool use, structured output, and multimodal prompts. It supports a 256k context window and configurable reasoning effort.","family":"mistral-medium","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.5,"output":7.5,"cache_read":0.75}},"mistralai/ministral-3b-2512":{"id":"mistralai/ministral-3b-2512","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.1,"output":0.1,"cache_read":0.05},"canonical_model_id":"mistral/ministral-3b-2512"},"TheDrummer/Cydonia-24B-v4.3":{"id":"TheDrummer/Cydonia-24B-v4.3","name":"The Drummer Cydonia 24B v4.3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-25","last_updated":"2025-12-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.12,"output":0.15,"cache_read":0.06}},"TheDrummer/Anubis-70B-v1.1":{"id":"TheDrummer/Anubis-70B-v1.1","name":"Anubis 70B v1.1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":16384},"cost":{"input":0.31,"output":0.31,"cache_read":0.155}},"TheDrummer/skyfall-36b-v2":{"id":"TheDrummer/skyfall-36b-v2","name":"TheDrummer Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":29491},"cost":{"input":0.55,"output":0.8,"cache_read":0.25},"canonical_model_id":"thedrummer/skyfall-36b-v2"},"TheDrummer/Rocinante-12B-v1.1":{"id":"TheDrummer/Rocinante-12B-v1.1","name":"Rocinante 12b","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.408,"output":0.595,"cache_read":0.204}},"TheDrummer/UnslopNemo-12B-v4.1":{"id":"TheDrummer/UnslopNemo-12B-v4.1","name":"UnslopNemo 12b v4","description":"Multimodal model for analyzing text, images, documents, and rich media","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"input":8192,"output":26214},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"TheDrummer/Anubis-70B-v1":{"id":"TheDrummer/Anubis-70B-v1","name":"Anubis 70B v1","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":16384},"cost":{"input":0.31,"output":0.31,"cache_read":0.155}},"TheDrummer/Artemis-v1.1":{"id":"TheDrummer/Artemis-v1.1","name":"TheDrummer/Artemis v1.1","description":"TheDrummer's Artemis v1.1 is a Gemma 4 31B fine-tune for creative writing, expressive dialogue, and roleplay, with optional thinking and a 262K context window.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-06","last_updated":"2026-09-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"TheDrummer/Cydonia-24B-v4.1":{"id":"TheDrummer/Cydonia-24B-v4.1","name":"The Drummer Cydonia 24B v4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.3,"output":0.5,"cache_read":0.15},"canonical_model_id":"thedrummer/cydonia-24b-v4.1"},"TheDrummer/skyfall-31b-v4.2":{"id":"TheDrummer/skyfall-31b-v4.2","name":"TheDrummer Skyfall 31B v4.2","description":"TheDrummer's Skyfall 31B v4.2 is a Magistral Small finetune for creative writing, roleplay, and dialogue. This text-only deployment uses NVFP4 weights and supports a 16K-token context window.","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":4096},"cost":{"input":0.12,"output":0.15,"cache_read":0.06}},"TheDrummer/orion-26b-a4b-v1.1":{"id":"TheDrummer/orion-26b-a4b-v1.1","name":"Orion 26B A4B v1.1","description":"TheDrummer's Gemma 4 26B A4B fine-tune for creative writing, roleplay, and expressive dialogue. Supports optional thinking with a 65,536-token context window and up to 8,192 output tokens. Text input only.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"TheDrummer/Magidonia-24B-v4.3":{"id":"TheDrummer/Magidonia-24B-v4.3","name":"The Drummer Magidonia 24B v4.3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-25","last_updated":"2025-12-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.1003,"output":0.1207,"cache_read":0.05015}},"TheDrummer/Cydonia-24B-v4":{"id":"TheDrummer/Cydonia-24B-v4","name":"The Drummer Cydonia 24B v4","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.2006,"output":0.2414,"cache_read":0.1003}},"TheDrummer/Cydonia-24B-v2":{"id":"TheDrummer/Cydonia-24B-v2","name":"The Drummer Cydonia 24B v2","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.1003,"output":0.1207,"cache_read":0.05015}},"featherless-ai/Qwerky-72B":{"id":"featherless-ai/Qwerky-72B","name":"Qwerky 72B","description":"General-purpose chat model for instruction following, writing, and analysis","family":"qwerky","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-03-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember 1","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"fireworks/ember-1"},"lightonai/LightOnOCR-2-1B":{"id":"lightonai/LightOnOCR-2-1B","name":"LightOnOCR 2","description":"LightOnOCR 2 hosted by IONOS in Berlin, Germany. Zero data retention.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.1785,"output":0.3465}},"z-ai/glm-5.3-flash-cybersecurity":{"id":"z-ai/glm-5.3-flash-cybersecurity","name":"GLM 5.3 Flash Cybersecurity","description":"GLM 5.3 Flash Cybersecurity is a cybersecurity-focused variant based on the uncensored model, with provider moderation for illegal activities. It supports always-on reasoning, image understanding, tool calling, and a 1,048,576-token context window.","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-19","last_updated":"2026-09-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.5,"cache_read":0.075}},"z-ai/glm-5v-turbo":{"id":"z-ai/glm-5v-turbo","name":"GLM 5V Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"input":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.7-flash:thinking":{"id":"z-ai/glm-4.7-flash:thinking","name":"GLM 4.7 Flash Thinking","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-4.5v":{"id":"z-ai/glm-4.5v","name":"GLM 4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-4.6":{"id":"z-ai/glm-4.6","name":"GLM 4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-5-original:thinking":{"id":"z-ai/glm-5-original:thinking","name":"GLM 5 Original Thinking","description":"GLM-5 original with extended thinking capabilities for complex reasoning.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.7-flash":{"id":"z-ai/glm-4.7-flash","name":"GLM 4.7 Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.07,"output":0.4,"cache_read":0.01},"canonical_model_id":"zhipuai/glm-4.7-flash"},"z-ai/glm-5.2":{"id":"z-ai/glm-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.189,"output":0.594,"cache_read":0.0351},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-4.7":{"id":"z-ai/glm-4.7","name":"GLM 4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.4004,"output":1.9292,"cache_read":0.08008},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/GLM-4.6-turbo:thinking":{"id":"z-ai/GLM-4.6-turbo:thinking","name":"GLM 4.6 Turbo (Thinking)","description":"GLM 4.6 Turbo with thinking mode enabled for enhanced reasoning; shows internal reasoning and supports long context.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":1,"output":3,"cache_read":0.5}},"z-ai/GLM-4.5-Air":{"id":"z-ai/GLM-4.5-Air","name":"GLM 4.5 Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":98304},"cost":{"input":0.12,"output":0.8,"cache_read":0.06},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-5.2:thinking":{"id":"z-ai/glm-5.2:thinking","name":"GLM 5.2 Thinking","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.189,"output":0.594,"cache_read":0.0351},"canonical_model_id":"zhipuai/glm-5.2"},"z-ai/glm-5.1":{"id":"z-ai/glm-5.1","name":"GLM 5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-5.1:thinking":{"id":"z-ai/glm-5.1:thinking","name":"GLM 5.1 Thinking","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"},"z-ai/glm-4.6-original":{"id":"z-ai/glm-4.6-original","name":"GLM 4.6 Original","description":"GLM-4.6, Zhipu's flagship text model with 256K context window and advanced reasoning capabilities. Direct via Z-AI (Zhipu).","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":65535},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.5":{"id":"z-ai/glm-4.5","name":"GLM 4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.3,"output":1.3,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-5v-turbo:thinking":{"id":"z-ai/glm-5v-turbo:thinking","name":"GLM 5V Turbo Thinking","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"input":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5v-turbo"},"z-ai/glm-4.7-original":{"id":"z-ai/glm-4.7-original","name":"GLM 4.7 Original","description":"GLM-4.7 is a next-gen GLM series text model with stronger reasoning, long-context chat, and reliable tool use. Routed directly via Z-AI (Zhipu).","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-5.3-flash-apex":{"id":"z-ai/glm-5.3-flash-apex","name":"Apex Flash 1","description":"Apex Flash 1 is Cantina Security and Yeta's cybersecurity-focused post-train of GLM 5.3 Flash for code investigation and tool use.","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.5,"cache_read":0.075}},"z-ai/glm-5":{"id":"z-ai/glm-5","name":"GLM 5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5.3":{"id":"z-ai/glm-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/GLM-4.5-Air:thinking":{"id":"z-ai/GLM-4.5-Air:thinking","name":"GLM 4.5 Air (Thinking)","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":98304},"cost":{"input":0.12,"output":0.8,"cache_read":0.06},"canonical_model_id":"zhipuai/glm-4.5-air"},"z-ai/glm-5-original":{"id":"z-ai/glm-5-original","name":"GLM 5 Original","description":"GLM-5 is Zhipu's latest flagship model with advanced reasoning and instruction following. Routed directly via Z-AI (Zhipu).","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-4.6:thinking":{"id":"z-ai/glm-4.6:thinking","name":"GLM 4.6 Thinking","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.5,"output":2,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-4.6"},"z-ai/glm-4.6v":{"id":"z-ai/glm-4.6v","name":"GLM 4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":24000},"cost":{"input":0.3,"output":0.9,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-4.5v:thinking":{"id":"z-ai/glm-4.5v:thinking","name":"GLM 4.5V Thinking","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":16384},"cost":{"input":0.6,"output":1.8,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-4.5v"},"z-ai/glm-5:thinking":{"id":"z-ai/glm-5:thinking","name":"GLM 5 Thinking","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":128000},"cost":{"input":0.6,"output":1.9,"cache_read":0.1},"canonical_model_id":"zhipuai/glm-5"},"z-ai/glm-5-turbo":{"id":"z-ai/glm-5-turbo","name":"GLM 5 Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":202800,"input":202800,"output":131072},"cost":{"input":1.2,"output":4,"cache_read":0.24},"canonical_model_id":"zhipuai/glm-5-turbo"},"z-ai/glm-5.3-flash":{"id":"z-ai/glm-5.3-flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.1,"output":0.3,"cache_read":0.025},"canonical_model_id":"zhipuai/glm-5.3-flash"},"z-ai/glm-latest":{"id":"z-ai/glm-latest","name":"GLM Latest","description":"Compatibility alias that routes to the newest thinking GLM model. Currently routes to GLM 5.2 Thinking.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2}},"z-ai/glm-5.3-uncensored":{"id":"z-ai/glm-5.3-uncensored","name":"GLM 5.3 Uncensored","description":"An experimental uncensored version of the full GLM 5.3 reasoning model for chat, creative writing, coding, and tool use. It may show unusual behavior or inconsistent response quality.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-26","last_updated":"2026-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.25,"output":2.25,"cache_read":0.3}},"z-ai/GLM-4.5:thinking":{"id":"z-ai/GLM-4.5:thinking","name":"GLM 4.5 (Thinking)","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.3,"output":1.3,"cache_read":0.15},"canonical_model_id":"zhipuai/glm-4.5"},"z-ai/glm-4.6v-original":{"id":"z-ai/glm-4.6v-original","name":"GLM 4.6V Original","description":"GLM-4.6V scales its context window to 128k tokens in training, and achieves SoTA performance in visual understanding among models of similar parameter scales. Integrates native Function Calling capabilities, bridging 'visual perception' and 'executable action' for multimodal agents. Direct via Z-AI (Zhipu).","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":24000},"cost":{"input":0.6,"output":0.9,"cache_read":0.3},"canonical_model_id":"zhipuai/glm-4.6v"},"z-ai/glm-5.3-flash-uncensored":{"id":"z-ai/glm-5.3-flash-uncensored","name":"GLM 5.3 Flash Uncensored","description":"GLM 5.3 Flash Uncensored is an uncensored fine-tune of the efficient 320B mixture-of-experts reasoning model, built for unrestricted chat, creative writing, coding, agentic work, tool use, and long-context tasks.","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.2,"output":0.8,"cache_read":0.07}},"z-ai/glm-5.3:thinking":{"id":"z-ai/glm-5.3:thinking","name":"GLM 5.3 Thinking","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.6,"output":1.1,"cache_read":0.2},"canonical_model_id":"zhipuai/glm-5.3"},"z-ai/glm-4.7:thinking":{"id":"z-ai/glm-4.7:thinking","name":"GLM 4.7 Thinking","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.4004,"output":1.9292,"cache_read":0.08008},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/glm-4.7-original:thinking":{"id":"z-ai/glm-4.7-original:thinking","name":"GLM 4.7 Original Thinking","description":"GLM-4.7 original with extended thinking capabilities for complex reasoning.","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":65535},"cost":{"input":0.6,"output":2.2,"cache_read":0.11},"canonical_model_id":"zhipuai/glm-4.7"},"z-ai/GLM-4.6-turbo":{"id":"z-ai/GLM-4.6-turbo","name":"GLM 4.6 Turbo","description":"Fast variant of GLM 4.6 for general chat, coding, and analysis with improved latency and strong reasoning.","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":1,"output":3,"cache_read":0.5}},"z-ai/glm-5.3-flash-apex-abliterated":{"id":"z-ai/glm-5.3-flash-apex-abliterated","name":"Apex Flash 1 Abliterated","description":"Apex Flash 1 Abliterated is an experimental derivative of Cantina Security and Yeta's Apex Flash 1 with modified refusal behavior.","family":"glm","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.5,"cache_read":0.075}},"google/gemini-3.1-pro-preview-low":{"id":"google/gemini-3.1-pro-preview-low","name":"Gemini 3.1 Pro (Preview Low)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.14,"output":0.4,"cache_read":0.05},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":0.1875},"canonical_model_id":"google/gemini-4-argon"},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-flash-lite-latest"},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash (Preview)","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.083333},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.8-flash"},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro (Preview)","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro (Preview Custom Tools)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview-customtools"},"google/gemini-3.5-flash-thinking":{"id":"google/gemini-3.5-flash-thinking","name":"Gemini 3.5 Flash Thinking","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.5,"output":9,"cache_read":0.15,"cache_write":0.083333},"canonical_model_id":"google/gemini-3.5-flash"},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.7-flash"},"google/gemini-3-flash-preview-thinking":{"id":"google/gemini-3-flash-preview-thinking","name":"Gemini 3 Flash Thinking","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":3,"cache_read":0.05},"canonical_model_id":"google/gemini-3-flash-preview"},"google/gemini-3.1-pro-preview-high":{"id":"google/gemini-3.1-pro-preview-high","name":"Gemini 3.1 Pro (Preview High)","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375},"canonical_model_id":"google/gemini-3.1-pro-preview"},"google/gemma-4-26b-a4b-it:thinking":{"id":"google/gemma-4-26b-a4b-it:thinking","name":"Gemma 4 26B A4B Thinking","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.0675,"output":0.225,"cache_read":0.0375},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/diffusiongemma":{"id":"google/diffusiongemma","name":"DiffusionGemma","description":"DiffusionGemma is a high-speed diffusion-based version of Gemma 4 26B A4B. It supports optional reasoning and a 262,144-token context window.","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","xhigh"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-07-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.05,"output":0.15,"cache_read":0.025},"canonical_model_id":"google/diffusiongemma-26b-a4b-it"},"google/gemma-4-26b-a4b-it-cybersecurity":{"id":"google/gemma-4-26b-a4b-it-cybersecurity","name":"Gemma 4 26B A4B Cybersecurity","description":"Gemma 4 26B A4B Cybersecurity is a cybersecurity-focused variant based on the uncensored model, with provider moderation for illegal activities. It supports optional reasoning, image understanding, tool calling, and a 262,144-token context window.","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-19","last_updated":"2026-09-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1056,"output":0.3344,"cache_read":0.0528}},"google/gemini-pro-latest":{"id":"google/gemini-pro-latest","name":"Gemini Pro Latest","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":0.375}},"google/gemma-4-31b-it:thinking":{"id":"google/gemma-4-31b-it:thinking","name":"Gemma 4 31B Thinking","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.1,"output":0.34,"cache_read":0.1},"canonical_model_id":"google/gemma-4-31b-it"},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-3.6-flash"},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.75,"output":3.75,"cache_read":0.075,"cache_write":0.041667},"canonical_model_id":"google/gemini-flash-latest"},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.0675,"output":0.225,"cache_read":0.0375},"canonical_model_id":"google/gemma-4-26b-a4b-it"},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.3,"output":2.5,"cache_read":0.03,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.5-flash-lite"},"google/gemma4-31b-splituntied":{"id":"google/gemma4-31b-splituntied","name":"Gemma 4 31B Split-Untied","description":"Blazed-Forge's Split-Untied is a text-only Gemma 4 31B community finetune with an untied BF16 output head, built for creative writing, roleplay, expressive dialogue, and tool use.","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.3,"cache_read":0.05}},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.25,"output":1.5,"cache_read":0.025,"cache_write":0.08333},"canonical_model_id":"google/gemini-3.1-flash-lite"},"shisa-ai/shisa-v2.1-llama3.3-70b":{"id":"shisa-ai/shisa-v2.1-llama3.3-70b","name":"Shisa V2.1 Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":4096},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"shisa-ai/shisa-v2-llama3.3-70b":{"id":"shisa-ai/shisa-v2-llama3.3-70b","name":"Shisa V2 Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"nvidia/nemotron-3.5-lightning:thinking":{"id":"nvidia/nemotron-3.5-lightning:thinking","name":"Nvidia Nemotron 3.5 Lightning Thinking","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nvidia Nemotron 3 Nano 30B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":235929},"cost":{"input":0.17,"output":0.68,"cache_read":0.085},"canonical_model_id":"nvidia/nemotron-3-nano-30b-a3b"},"nvidia/Llama-3.3-Nemotron-Super-49B-v1":{"id":"nvidia/Llama-3.3-Nemotron-Super-49B-v1","name":"Nvidia Nemotron Super 49B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-03-18","last_updated":"2025-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.15,"output":0.15,"cache_read":0.075},"canonical_model_id":"nvidia/llama-3.3-nemotron-super-49b-v1"},"nvidia/nemotron-3-super-120b-a12b:thinking":{"id":"nvidia/nemotron-3-super-120b-a12b:thinking","name":"Nvidia Nemotron 3 Super 120B Thinking","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.05,"output":0.25,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.05,"output":0.15,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3.5-content-safety"},"nvidia/nemotron-3-ultra-550b-a55b:thinking":{"id":"nvidia/nemotron-3-ultra-550b-a55b:thinking","name":"Nvidia Nemotron 3 Ultra 550B Thinking","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nvidia Nemotron 3 Super 120B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.05,"output":0.25,"cache_read":0.025},"canonical_model_id":"nvidia/nemotron-3-super-120b-a12b"},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nvidia Nemotron 3 Ultra 550B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"nvidia/nemotron-3-ultra-550b-a55b"},"nvidia/Llama-3.1-Nemotron-70B-Instruct-HF":{"id":"nvidia/Llama-3.1-Nemotron-70B-Instruct-HF","name":"Nvidia Nemotron 70b","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.357,"output":0.408,"cache_read":0.1785},"canonical_model_id":"nvidia/llama-3.1-nemotron-70b-instruct"},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nvidia Nemotron 3.5 Lightning","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.05,"output":0.2,"cache_read":0.01},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"nvidia/nemotron-3.5-super-vl":{"id":"nvidia/nemotron-3.5-super-vl","name":"Nvidia Nemotron 3.5 Super VL","description":"NVIDIA’s open-weight 120B model (12B active) for reasoning, coding, tool use, document and GUI understanding, and video understanding. Text, images and video in; text out. Served in NVFP4 with a 262,144-token shared context window. NanoGPT currently limits responses to 16,384 tokens while separate hosted output limits remain unpublished. OpenMDW 1.1 license.","family":"nemotron","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.15,"output":0.5,"cache_read":0.02},"canonical_model_id":"nvidia/nemotron-3.5-super-vl-120b-a12b"},"inception/mercury-2.5-preview":{"id":"inception/mercury-2.5-preview","name":"Mercury 2.5 Preview","description":"Mercury 2.5 Preview is Inception's latest and most intelligent diffusion language model. Instead of generating tokens strictly one at a time, it produces and refines multiple tokens in parallel, reaching up to 1,107 tokens per second on standard GPUs. It delivers a 10+ point intelligence gain over Mercury 2, with tunable reasoning, parallel tool calls, schema-aligned JSON output, and a 260K context window. It is built for latency-sensitive production work such as search agents, voice pipelines, customer support, rapid coding iteration, and coding subagents.","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"input":260000,"output":65536},"cost":{"input":0.04,"output":0.15,"cache_read":0.004}},"inception/mercury-3":{"id":"inception/mercury-3","name":"Mercury 3","description":"Inception's diffusion reasoning model for chat, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.1,"output":0.35,"cache_read":0.01},"canonical_model_id":"inception/mercury-3"},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo V2.6 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro"},"xiaomi/mimo-v2.6-flash-abliterated":{"id":"xiaomi/mimo-v2.6-flash-abliterated","name":"MiMo V2.6 Flash Abliterated","description":"A MiMo V2.6 Flash finetune with refusal-direction ablation, image input, a 1M-token context window, separate reasoning output, and tool calling.","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":1.5,"cache_read":0.275}},"xiaomi/mimo-v2.6-flash-uncensored":{"id":"xiaomi/mimo-v2.6-flash-uncensored","name":"MiMo V2.6 Flash Uncensored","description":"A text-only, lower-refusal MiMo V2.6 Flash finetune served by Heabsy with a 1M-token context window, separate reasoning output, and tool calling.","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":1.5,"cache_read":0.1995}},"xiaomi/mimo-v2.5:thinking":{"id":"xiaomi/mimo-v2.5:thinking","name":"MiMo V2.5 Thinking","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5"},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo V2.6 Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.6-flash"},"xiaomi/mimo-v2.6-flash-uncensored:thinking":{"id":"xiaomi/mimo-v2.6-flash-uncensored:thinking","name":"MiMo V2.6 Flash Uncensored Thinking","description":"MiMo V2.6 Flash Uncensored with maximum thinking enabled. Supports a 1M-token context window and tool calling.","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-26","last_updated":"2026-09-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.5,"output":1.5,"cache_read":0.1995}},"xiaomi/mimo-v2.5-pro:thinking":{"id":"xiaomi/mimo-v2.5-pro:thinking","name":"MiMo V2.5 Pro Thinking","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo V2.6 Pro UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":4.35,"output":8.7,"cache_read":0.036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.6-pro-ultraspeed"},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo V2.5 Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.435,"output":0.87,"cache_read":0.0036,"cache_write":0},"canonical_model_id":"xiaomi/mimo-v2.5-pro"},"mlabonne/NeuralDaredevil-8B-abliterated":{"id":"mlabonne/NeuralDaredevil-8B-abliterated","name":"Neural Daredevil 8B abliterated","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":0.44,"output":0.44,"cache_read":0.22}},"xingchen-agi/xing4.0-29b-a4b":{"id":"xingchen-agi/xing4.0-29b-a4b","name":"Xing 4.0 29B A4B","description":"China Telecom’s Xing 4.0 is an Apache 2.0 open-weight mixture-of-experts model with 29B total and 4B active parameters, designed for coding, reasoning and tool use. Supports text input, structured output and a shared 262,144-token input/output context. This route does not use a TEE.","family":"agi","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-24","last_updated":"2026-10-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.22,"output":0.9,"cache_read":0.011}},"elorian/glimpse":{"id":"elorian/glimpse","name":"Elorian Glimpse","description":"Glimpse is Elorian's visual reasoning model for understanding images and video and answering in text. It can analyze charts, count objects, follow events over time, and reason about spatial relationships and physical cause and effect. It supports tool calling and structured output across a shared 262,144-token context window, with up to 256,000 output tokens within the remaining context.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":256000},"cost":{"input":0.63,"output":3.15,"cache_read":0.063}},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":1000000},"cost":{"input":1,"output":2.7,"cache_read":0.05},"canonical_model_id":"stepfun/step-5-preview"},"stepfun/step-3.7-flash:thinking":{"id":"stepfun/step-3.7-flash:thinking","name":"Step 3.7 Flash Thinking","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":256000},"cost":{"input":0.2,"output":1.15,"cache_read":0.04},"canonical_model_id":"stepfun/step-3.7-flash"},"VongolaChouko/Starcannon-Unleashed-12B-v1.0":{"id":"VongolaChouko/Starcannon-Unleashed-12B-v1.0","name":"Mistral Nemo Starcannon 12b v1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Steelskull/L3.3-Cu-Mai-R1-70b":{"id":"Steelskull/L3.3-Cu-Mai-R1-70b","name":"Llama 3.3 70B Cu Mai","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Steelskull/L3.3-MS-Nevoria-70b":{"id":"Steelskull/L3.3-MS-Nevoria-70b","name":"Steelskull Nevoria 70b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Steelskull/L3.3-Electra-R1-70b":{"id":"Steelskull/L3.3-Electra-R1-70b","name":"Steelskull Electra R1 70b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.69989,"output":0.69989,"cache_read":0.349945}},"Steelskull/L3.3-Nevoria-R1-70b":{"id":"Steelskull/L3.3-Nevoria-R1-70b","name":"Steelskull Nevoria R1 70b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"nimbz/schattenblume-31b":{"id":"nimbz/schattenblume-31b","name":"Schattenblume 31B","description":"Nimbz's Gemma 4 31B merge for character roleplay, expressive prose, and creative dialogue. It combines Scotoma 2, Giftige-Blume v1, and MeroMero v2. This text-only deployment supports a 65,536-token context window and up to 8,192 output tokens.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"nimbz/gemma-4-giftige-blume-31b-v2":{"id":"nimbz/gemma-4-giftige-blume-31b-v2","name":"Gemma 4 Giftige-Blume V2 31B","description":"Nimbz's second Giftige-Blume merge focuses on character roleplay, creative prose, and stricter prompt adherence. It combines storymaxxed, Equinox, Mero-Artemis, Musica, Ortenzya, Gemopus, and Sprinkle. This text-only Gemma 4 31B deployment supports a 65,536-token context window and up to 8,192 output tokens.","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"MiniMaxAI/MiniMax-M1-80k":{"id":"MiniMaxAI/MiniMax-M1-80k","name":"MiniMax M1 80K","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":0.6052,"output":2.4225,"cache_read":0.3026},"canonical_model_id":"minimax/minimax-m1-80k"},"perceptron/perceptron-mk1":{"id":"perceptron/perceptron-mk1","name":"Perceptron Mk1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.15,"output":1.5,"cache_read":0.075},"canonical_model_id":"perceptron/perceptron-mk1"},"perceptron/perceptron-mk1.5":{"id":"perceptron/perceptron-mk1.5","name":"Perceptron Mk1.5","description":"Perceptron Mk1.5 is a multimodal reasoning model for visual analysis and physical agents. It accepts text, images, video, and audio, supports tool calling and structured output, and can return spatial annotations such as points, boxes, polygons, and tracks.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":36864,"input":36864,"output":8192},"cost":{"input":0.15,"output":1.5}},"NeverSleep/Lumimaid-v0.2-70B":{"id":"NeverSleep/Lumimaid-v0.2-70B","name":"Lumimaid v0.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":1,"output":1.5,"cache_read":0.5}},"abacusai/Dracarys-72B-Instruct":{"id":"abacusai/Dracarys-72B-Instruct","name":"Llama 3.1 70B Dracarys 2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-08-02","last_updated":"2025-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"thinkingmachines/Inkling-Small":{"id":"thinkingmachines/Inkling-Small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling:thinking":{"id":"thinkingmachines/inkling:thinking","name":"Inkling Thinking","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"input":1048000,"output":32768},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"thinkingmachines/Inkling-Small:thinking":{"id":"thinkingmachines/Inkling-Small:thinking","name":"Inkling Small Thinking","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.45,"output":1.2,"cache_read":0.1},"canonical_model_id":"thinkingmachines/inkling-small"},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048000,"input":1048000,"output":32768},"cost":{"input":1,"output":4.05,"cache_read":0.17},"canonical_model_id":"thinkingmachines/inkling"},"unbiased/pareto-26.9":{"id":"unbiased/pareto-26.9","name":"Pareto 26.9","description":"Pareto 26.9 is an experimental multimodal composite model from Unbiased AI for research, coding, agentic workflows, and general-purpose tasks. It runs several frontier and open models for each request and returns a single selected or synthesized answer. This experimental release supports text and image input, tool calling, a 262K-token context window, and up to 131K output tokens. Its underlying model mix and behavior may change as it is evaluated. Request content may be retained for abuse prevention and security monitoring for up to 30 days, and is not used for model training without explicit consent.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":2.5,"output":7.5,"cache_read":0.25}},"unbiased/pareto-26.10":{"id":"unbiased/pareto-26.10","name":"Pareto 26.10","description":"Pareto 26.10 is Unbiased AI's multimodal composite model for research, coding, and agentic workflows. It accepts text and images, supports tool calling, and provides a 1M-token context window with up to 131K output tokens. This is the stable release of Pareto 26.10. Tool selection is automatic; forced tool selection, structured output, and adjustable reasoning controls are not supported.","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.8,"output":3.2,"cache_read":0.03}},"aisingapore/Gemma-SEA-LION-v4-27B-IT":{"id":"aisingapore/Gemma-SEA-LION-v4-27B-IT","name":"Gemma SEA-LION v4 27B","description":"Gemma 3 27B tuned by AI Singapore for Southeast Asian languages and instruction following","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":128000},"cost":{"input":0.24,"output":0.48},"canonical_model_id":"aisingapore/gemma-sea-lion-v4-27b-it"},"reflection/beam-1.0":{"id":"reflection/beam-1.0","name":"Beam 1.0 501B A23B","description":"Reflection's Apache 2.0 open-weight text model with 501.3B total and 23.4B active parameters, built for coding, reasoning and tool use. It supports a shared 1,048,576-token context window, up to 131,072 output tokens, JSON mode and structured output. Reasoning is always enabled with low, medium, high, xhigh and max effort; NanoGPT defaults to max. Knowledge cutoff: June 30, 2026.","family":"beam","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.5,"output":1.5,"cache_read":0.05},"canonical_model_id":"reflection/beam-1.0-501b-a23b"},"LLM360/K2-Think":{"id":"LLM360/K2-Think","name":"K2-Think","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":0.17,"output":0.68,"cache_read":0.085}},"deepcogito/cogito-v1-preview-qwen-32B":{"id":"deepcogito/cogito-v1-preview-qwen-32B","name":"Cogito v1 Preview Qwen 32B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-05-10","last_updated":"2025-05-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":1.8,"output":1.8,"cache_read":0.9}},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Tencent Hy4 Preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"tencent/hy3":{"id":"tencent/hy3","name":"Tencent Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":128000},"cost":{"input":0.066,"output":0.26,"cache_read":0.029},"canonical_model_id":"tencent/hy3"},"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":80000},"cost":{"input":0.25,"output":0.9,"cache_read":0.125},"canonical_model_id":"arcee-ai/trinity-large-thinking"},"pamanseau/OpenReasoning-Nemotron-32B":{"id":"pamanseau/OpenReasoning-Nemotron-32B","name":"OpenReasoning Nemotron 32B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05}},"meta-llama/llama-3.3-70b-instruct":{"id":"meta-llama/llama-3.3-70b-instruct","name":"Llama 3.3 70b Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.05,"output":0.23,"cache_read":0.025},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"meta-llama/llama-4-maverick":{"id":"meta-llama/llama-4-maverick","name":"Llama 4 Maverick","description":"Open multimodal Llama model for strong reasoning and fast responses","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.1875,"output":0.6525,"cache_read":0.05},"canonical_model_id":"meta/llama-4-maverick-17b-instruct"},"meta-llama/llama-3.1-8b-instruct":{"id":"meta-llama/llama-3.1-8b-instruct","name":"Llama 3.1 8b Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.02,"output":0.04,"cache_read":0.0272},"canonical_model_id":"meta/llama-3.1-8b-instruct"},"meta-llama/llama-4-scout":{"id":"meta-llama/llama-4-scout","name":"Llama 4 Scout","description":"Open multimodal Llama model for long-context analysis and efficient agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":328000,"input":328000,"output":65536},"cost":{"input":0.085,"output":0.46,"cache_read":0.0425},"canonical_model_id":"meta/llama-4-scout-17b-instruct"},"meta-llama/llama-3.2-3b-instruct":{"id":"meta-llama/llama-3.2-3b-instruct","name":"Llama 3.2 3b Instruct","description":"Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":8192},"cost":{"input":0.0306,"output":0.0493,"cache_read":0.0153},"canonical_model_id":"meta/llama-3.2-3b-instruct"},"ornith-ai/ornith-1.5-35b-a3b:thinking":{"id":"ornith-ai/ornith-1.5-35b-a3b:thinking","name":"Ornith 1.5 35B Thinking","description":"Ornith 1.5 35B A3B is an open-weight mixture-of-experts model for agentic coding, reasoning, tool use, image understanding, and long-context work. This variant enables thinking by default.","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.05}},"ornith-ai/ornith-1.5-35b-a3b":{"id":"ornith-ai/ornith-1.5-35b-a3b","name":"Ornith 1.5 35B","description":"Ornith 1.5 35B A3B is an open-weight mixture-of-experts model for agentic coding, tool use, image understanding, and long-context work. This variant disables thinking for faster direct responses.","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.05}},"ornith-ai/ornith-1.5-122b":{"id":"ornith-ai/ornith-1.5-122b","name":"Ornith 1.5 122B","description":"Ornith 1.5 122B is an MIT-licensed open-weight model for coding, reasoning, tool use and image understanding. It accepts text and images across a shared 262,144-token context, with up to 32,768 output tokens. Reasoning is enabled by default; JSON object output and streaming are supported.","family":"ornith","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-20","last_updated":"2026-10-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.2,"output":0.8,"cache_read":0.05}},"deepseek/deepseek-v4.1-flash-1017:thinking":{"id":"deepseek/deepseek-v4.1-flash-1017:thinking","name":"DeepSeek V4.1 Flash 1017 Thinking","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":393216},"cost":{"input":0.1,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-latest":{"id":"deepseek/deepseek-latest","name":"DeepSeek Latest","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.132,"output":0.396,"cache_read":0.0042}},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":0.0537,"output":0.2148,"cache_read":0.001074},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4-flash:thinking":{"id":"deepseek/deepseek-v4-flash:thinking","name":"DeepSeek V4 Flash (Thinking)","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":393216},"cost":{"input":0.1,"output":0.6,"cache_read":0.003},"canonical_model_id":"deepseek/deepseek-v4.1-flash-1017"},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.7,"output":1.4,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4.1-flash:thinking":{"id":"deepseek/deepseek-v4.1-flash:thinking","name":"DeepSeek V4.1 Flash Thinking","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":384000},"cost":{"input":0.0537,"output":0.2148,"cache_read":0.001074},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"deepseek/deepseek-v4.1-flash-uncensored:thinking":{"id":"deepseek/deepseek-v4.1-flash-uncensored:thinking","name":"DeepSeek V4.1 Flash Uncensored Thinking","description":"A lower-refusal finetune of DeepSeek V4.1 Flash with text input, optional reasoning, and tool calling. Supports a context window of up to 1M tokens.","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.33,"output":1.32,"cache_read":0.00825}},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.132,"output":0.396,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.04444,"output":0.13332,"cache_read":0.001414},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-pro:thinking":{"id":"deepseek/deepseek-v4-pro:thinking","name":"DeepSeek V4 Pro (Thinking)","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.7,"output":1.4,"cache_read":0.07},"canonical_model_id":"deepseek/deepseek-v4-pro"},"deepseek/deepseek-v4.1-flash-uncensored":{"id":"deepseek/deepseek-v4.1-flash-uncensored","name":"DeepSeek V4.1 Flash Uncensored","description":"A lower-refusal finetune of DeepSeek V4.1 Flash with text input, optional reasoning, and tool calling. Supports a context window of up to 1M tokens.","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":0.33,"output":1.32,"cache_read":0.00825}},"deepseek/deepseek-v4-flash-0423":{"id":"deepseek/deepseek-v4-flash-0423","name":"DeepSeek V4 Flash 0423","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-v4-flash-0731:thinking":{"id":"deepseek/deepseek-v4-flash-0731:thinking","name":"DeepSeek V4 Flash 0731 (Thinking)","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.04444,"output":0.13332,"cache_read":0.001414},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"deepseek/deepseek-v4-flash-0423:thinking":{"id":"deepseek/deepseek-v4-flash-0423:thinking","name":"DeepSeek V4 Flash 0423 (Thinking)","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028},"canonical_model_id":"deepseek/deepseek-v4-flash-0423"},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":163000,"input":163000,"output":65536},"cost":{"input":0.2088,"output":0.3096,"cache_read":0.0216},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-flash-latest":{"id":"deepseek/deepseek-v4-flash-latest","name":"DeepSeek V4 Flash Latest","description":"Compatibility alias that routes to the newest dated DeepSeek V4 Flash release. Currently routes to DeepSeek V4 Flash 0731. ⚠️ This route goes directly to DeepSeek, so privacy and logging guarantees are limited.","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":false,"release_date":"2026-08-02","last_updated":"2026-08-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":393216},"cost":{"input":0.1,"output":0.6,"cache_read":0.003}},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.14,"output":0.28,"cache_read":0.0028},"canonical_model_id":"deepseek/deepseek-v4-flash"},"deepseek/deepseek-v3.2:thinking":{"id":"deepseek/deepseek-v3.2:thinking","name":"DeepSeek V3.2 Thinking","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":163000,"input":163000,"output":65536},"cost":{"input":0.2088,"output":0.3096,"cache_read":0.0216},"canonical_model_id":"deepseek/deepseek-v3.2"},"deepseek/deepseek-v4-pro-0813:thinking":{"id":"deepseek/deepseek-v4-pro-0813:thinking","name":"DeepSeek V4 Pro 0813 Thinking","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.132,"output":0.396,"cache_read":0.0042},"canonical_model_id":"deepseek/deepseek-v4-pro-0813"},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.2156,"output":0.6468,"cache_read":0.00686},"canonical_model_id":"deepseek/deepseek-v4-flash-vision-exp"},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"moonshotai/kimi-latest":{"id":"moonshotai/kimi-latest","name":"Kimi Latest","description":"Kimi multimodal agent model for visual understanding, coding, and planning","family":"kimi","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":1.2,"output":5.5,"cache_read":0.2}},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":1048576},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2},"canonical_model_id":"moonshotai/kimi-k3.1-pro"},"moonshotai/Kimi-K2-Instruct-0905":{"id":"moonshotai/Kimi-K2-Instruct-0905","name":"Kimi K2 0905","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":100352},"cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-0905"},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.375,"output":2,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2.6:thinking":{"id":"moonshotai/kimi-k2.6:thinking","name":"Kimi K2.6 Thinking","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.44935,"output":1.892,"cache_read":0.07568},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":1.2,"output":5.5,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k3"},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.44935,"output":1.892,"cache_read":0.07568},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/kimi-k2-instruct":{"id":"moonshotai/kimi-k2-instruct","name":"Kimi K2 Instruct","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":8192},"cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/kimi-k2.5:thinking":{"id":"moonshotai/kimi-k2.5:thinking","name":"Kimi K2.5 Thinking","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.375,"output":2,"cache_read":0.07},"canonical_model_id":"moonshotai/kimi-k2.5"},"moonshotai/kimi-k2-instruct-0711":{"id":"moonshotai/kimi-k2-instruct-0711","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.4,"output":1.8,"cache_read":0.2},"canonical_model_id":"moonshotai/kimi-k2-instruct"},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":98304},"cost":{"input":0.6,"output":2.5,"cache_read":0.15},"canonical_model_id":"moonshotai/kimi-k2-thinking"},"moonshotai/kimi-k3.1-pro-highspeed":{"id":"moonshotai/kimi-k3.1-pro-highspeed","name":"Kimi K3.1 Pro Highspeed","description":"Kimi K3.1 Pro Highspeed is the faster variant of K3.1 Pro, a multimodal reasoning model for coding, 3D creation, document analysis and multi-step tool use. It supports a 1M-token context window and text, image and video inputs. Reasoning is always enabled, with low, high and max effort levels.","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-24","last_updated":"2026-10-24","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":1048576},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":4}},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code High-Speed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":1.9,"output":8,"cache_read":0.32},"canonical_model_id":"moonshotai/kimi-k2.7-code-highspeed"},"soob3123/Veiled-Calla-12B":{"id":"soob3123/Veiled-Calla-12B","name":"Veiled Calla 12B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-04-13","last_updated":"2025-04-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"soob3123/amoral-gemma3-27B-v2":{"id":"soob3123/amoral-gemma3-27B-v2","name":"Amoral Gemma3 27B v2","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-05-23","last_updated":"2025-05-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.3,"output":0.3,"cache_read":0.15}},"meta/muse-spark-1.3-contributor":{"id":"meta/muse-spark-1.3-contributor","name":"Muse Spark 1.3 Contributor","description":"Meta's Muse Spark 1.3 Contributor is a frontier multimodal reasoning model for long-horizon coding and agentic workflows, with strong gains in computer use, browsing, professional tool use, codebase understanding, and million-token retrieval. It accepts text, images, audio, video, and files, supports tool calling and structured output, and always reasons before answering. Prompts and outputs may be used by Meta for training and to improve its products.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.35,"output":1.5,"cache_read":0.04},"canonical_model_id":"meta/muse-glimmer-30b"},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.1"},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":943718},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.3"},"meta/muse-spark-1.2-contributor":{"id":"meta/muse-spark-1.2-contributor","name":"Muse Spark 1.2 Contributor (Data Used for Training)","description":"A much cheaper opt-in version of Muse Spark 1.2 with the same multimodal coding and agentic capabilities. Prompts and outputs sent to this Contributor model may be used by Meta for training and to improve its products; use the standard Muse Spark 1.2 model if you do not want your data used for training.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":0.1,"output":0.2,"cache_read":0.002}},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-1.2"},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.25,"output":4.25,"cache_read":0.15},"canonical_model_id":"meta/muse-spark-open"},"nanogpt/coding-router:low":{"id":"nanogpt/coding-router:low","name":"Coding Router Low","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028}},"nanogpt/coding-router:medium":{"id":"nanogpt/coding-router:medium","name":"Coding Router Medium","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.14,"output":0.28,"cache_read":0.028}},"nanogpt/coding-router:max":{"id":"nanogpt/coding-router:max","name":"Coding Router Max","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5}},"nanogpt/coding-router":{"id":"nanogpt/coding-router","name":"Coding Router","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":1.1,"output":2.2,"cache_read":0.11}},"nanogpt/coding-router:high":{"id":"nanogpt/coding-router:high","name":"Coding Router High","description":"Automatic model router for matching prompts to suitable backends and budgets","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":1.1,"output":2.2,"cache_read":0.11}},"nbeerbower/gemma4-gutenberg-31b":{"id":"nbeerbower/gemma4-gutenberg-31b","name":"Gemma 4 Gutenberg 31B","description":"nbeerbower's Gemma 4 31B finetune for literary fiction, character interiority, and controlled narrative pacing. Its text decoder was trained with ORPO on 14,816 Athanorlite-DPO preference pairs. This text-only deployment supports a 65,536-token context window and up to 8,192 output tokens.","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"baseten/Kimi-K2-Instruct-FP4":{"id":"baseten/Kimi-K2-Instruct-FP4","name":"Kimi K2 0711 Instruct FP4","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":131072},"status":"deprecated","cost":{"input":0.4,"output":1.8,"cache_read":0.2}},"unsloth/gemma-3-27b-it":{"id":"unsloth/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":96000},"cost":{"input":0.2992,"output":0.2992,"cache_read":0.1496},"canonical_model_id":"google/gemma-3-27b-it"},"unsloth/gemma-3-12b-it":{"id":"unsloth/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.272,"output":0.272,"cache_read":0.136},"canonical_model_id":"google/gemma-3-12b-it"},"unsloth/gemma-3-4b-it":{"id":"unsloth/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003},"canonical_model_id":"google/gemma-3-4b-it"},"LatitudeGames/Wayfarer-Large-70B-Llama-3.3":{"id":"LatitudeGames/Wayfarer-Large-70B-Llama-3.3","name":"Llama 3.3 70B Wayfarer","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-02-20","last_updated":"2025-02-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"upstage/solar-pro-3":{"id":"upstage/solar-pro-3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"upstage/solar-pro3"},"upstage/solar-pro4:thinking":{"id":"upstage/solar-pro4:thinking","name":"Solar Pro 4 Thinking","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-mini4":{"id":"upstage/solar-mini4","name":"Solar Mini 4","description":"Upstage's compact, cost-efficient model for agentic use","family":"solar-mini","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.005},"canonical_model_id":"upstage/solar-mini4"},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.03,"output":0.12,"cache_read":0.006},"canonical_model_id":"upstage/solar-pro4"},"upstage/solar-mini4:thinking":{"id":"upstage/solar-mini4:thinking","name":"Solar Mini 4 Thinking","description":"Upstage's compact, cost-efficient model for agentic use","family":"solar-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"input":524288,"output":131072},"cost":{"input":0.05,"output":0.2,"cache_read":0.005},"canonical_model_id":"upstage/solar-mini4"},"TEE/kimi-k2.7-code":{"id":"TEE/kimi-k2.7-code","name":"Kimi K2.7 Code TEE","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.95,"output":4,"cache_read":0.19},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"TEE/gemma-4-31b-it":{"id":"TEE/gemma-4-31b-it","name":"Gemma 4 31B IT TEE","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.15,"output":0.46,"cache_read":0.075},"canonical_model_id":"google/gemma-4-31b-it"},"TEE/gpt-oss-120b":{"id":"TEE/gpt-oss-120b","name":"GPT-OSS 120B TEE","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":2,"output":2,"cache_read":2},"canonical_model_id":"openai/gpt-oss-120b"},"TEE/glm-5.2":{"id":"TEE/glm-5.2","name":"GLM 5.2 TEE","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.7},"canonical_model_id":"zhipuai/glm-5.2"},"TEE/deepseek-v4.1-flash":{"id":"TEE/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash TEE","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":384000},"cost":{"input":0.65,"output":1.45,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"TEE/qwen2.5-vl-72b-instruct":{"id":"TEE/qwen2.5-vl-72b-instruct","name":"Qwen2.5 VL 72B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35},"canonical_model_id":"alibaba/qwen2-5-vl-72b-instruct"},"TEE/glm-5.2:thinking":{"id":"TEE/glm-5.2:thinking","name":"GLM 5.2 Thinking TEE","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.7},"canonical_model_id":"zhipuai/glm-5.2"},"TEE/glm-5.1":{"id":"TEE/glm-5.1","name":"GLM 5.1 TEE","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"input":202752,"output":65535},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"},"TEE/gemma-4-26b-a4b-uncensored":{"id":"TEE/gemma-4-26b-a4b-uncensored","name":"Gemma 4 26B A4B Uncensored TEE","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2026-05-23","last_updated":"2026-05-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":65536},"cost":{"input":0.15,"output":0.7,"cache_read":0.075}},"TEE/qwen3.6-27b":{"id":"TEE/qwen3.6-27b","name":"Qwen3.6 27B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.32,"output":2.7,"cache_read":0.16},"canonical_model_id":"alibaba/qwen3.6-27b"},"TEE/kimi-k3":{"id":"TEE/kimi-k3","name":"Kimi K3 TEE","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65535},"cost":{"input":3,"output":15,"cache_read":1.5},"canonical_model_id":"moonshotai/kimi-k3"},"TEE/kimi-k2.6":{"id":"TEE/kimi-k2.6","name":"Kimi K2.6 TEE","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":1.5,"output":5.25,"cache_read":0.375},"canonical_model_id":"moonshotai/kimi-k2.6"},"TEE/gemma4-31b:thinking":{"id":"TEE/gemma4-31b:thinking","name":"Gemma 4 31B Thinking TEE","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.4,"output":1,"cache_read":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"TEE/qwen3.5-397b-a17b":{"id":"TEE/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B TEE","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.55,"output":3.5,"cache_read":0.275},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"TEE/glm-5.1-thinking":{"id":"TEE/glm-5.1-thinking","name":"GLM 5.1 Thinking TEE","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":202752,"input":202752,"output":65535},"cost":{"input":0.7,"output":2.5,"cache_read":0.18},"canonical_model_id":"zhipuai/glm-5.1"},"TEE/gemma4-31b":{"id":"TEE/gemma4-31b","name":"Gemma 4 31B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.4,"output":1,"cache_read":0.4},"canonical_model_id":"google/gemma-4-31b-it"},"TEE/llama3-3-70b":{"id":"TEE/llama3-3-70b","name":"Llama 3.3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":1.75,"output":2.75,"cache_read":1.75},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"TEE/glm-5.3":{"id":"TEE/glm-5.3","name":"GLM 5.3 TEE","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":1.4,"output":4.4,"cache_read":0.26},"canonical_model_id":"zhipuai/glm-5.3"},"TEE/qwen3.6-35b-a3b":{"id":"TEE/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B TEE","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.2,"output":1.27,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"TEE/deepseek-v3.2":{"id":"TEE/deepseek-v3.2","name":"DeepSeek V3.2 TEE","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":164000,"input":164000,"output":65536},"cost":{"input":0.5,"output":1,"cache_read":0.25},"canonical_model_id":"deepseek/deepseek-v3.2"},"TEE/glm-5.3-flash":{"id":"TEE/glm-5.3-flash","name":"GLM 5.3 Flash TEE","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.15,"output":0.5,"cache_read":0.03},"canonical_model_id":"zhipuai/glm-5.3-flash"},"TEE/qwen3.5-27b":{"id":"TEE/qwen3.5-27b","name":"Qwen3.5 27B TEE","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.3,"output":2.4,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.5-27b"},"TEE/qwen3.8-27b":{"id":"TEE/qwen3.8-27b","name":"Qwen3.8 27B TEE","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":262144},"cost":{"input":0.4,"output":3,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3.8-27b"},"TEE/nemotron-3.5-lightning":{"id":"TEE/nemotron-3.5-lightning","name":"Nvidia Nemotron 3.5 Lightning TEE","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.08,"output":0.2,"cache_read":0.04},"canonical_model_id":"nvidia/nemotron-3.5-lightning"},"GalrionSoftworks/MN-LooseCannon-12B-v1":{"id":"GalrionSoftworks/MN-LooseCannon-12B-v1","name":"MN-LooseCannon-12B-v1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"liquid/lfm-2.5-24b-a2b":{"id":"liquid/lfm-2.5-24b-a2b","name":"LFM2.5 24B A2B","description":"Liquid AI's text-only mixture-of-experts model for reasoning and tool-using agents, with 24B total and 2.45B active parameters. It always thinks before answering and supports structured output. Pair it with retrieval for knowledge-heavy tasks. Open weights use the LFM Open License v1.0, with a $10M annual-revenue commercial threshold; hosted API use is subject to separate service terms that permit retention and use of prompts and responses for model training.","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-10-14","last_updated":"2026-10-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.05,"output":0.7,"cache_read":0.025}},"liquid/lfm-2.5-2.6b":{"id":"liquid/lfm-2.5-2.6b","name":"LFM2.5 2.6B","description":"Liquid AI's compact 2.6B reasoning model for agent workflows, data extraction, RAG, and long-context processing. It supports tool calling and structured output, but Liquid advises against using it for agentic coding. Warning: prompts and responses may be logged and used for model training or service improvement; do not send sensitive data.","family":"liquid","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":0.1,"output":0.2,"cache_read":0.05}},"ibm-granite/granite-4.2-8b":{"id":"ibm-granite/granite-4.2-8b","name":"Granite 4.2 8B","description":"IBM Granite 4.2 8B is an Apache 2.0-licensed dense model with native step-by-step reasoning and specialized training for agentic work. It can plan before acting, sequence tools, navigate codebases, work in terminals, and verify results across coding, search, mathematics, science, and complex instruction-following tasks.","family":"granite","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":117964},"cost":{"input":0.1,"output":0.15,"cache_read":0.05},"canonical_model_id":"ibm/granite-4.2-8b"},"aion-labs/aion-3.0-mini":{"id":"aion-labs/aion-3.0-mini","name":"AionLabs: Aion 3.0 Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18},"canonical_model_id":"aion-labs/aion-3.0-mini"},"aion-labs/aion-3.5-mini":{"id":"aion-labs/aion-3.5-mini","name":"AionLabs: Aion 3.5 Mini","description":"A GLM-family collaborative generation model tuned for immersive roleplay and storytelling, with stronger narrative structure, tension, conflict, and nuanced mature themes.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.7,"output":1.4,"cache_read":0.18}},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"Llama 3.1 8b (uncensored)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.8,"output":1.6,"cache_read":0.4},"canonical_model_id":"aion-labs/aion-rp-llama-3.1-8b"},"aion-labs/aion-3.0":{"id":"aion-labs/aion-3.0","name":"AionLabs: Aion 3.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75},"canonical_model_id":"aion-labs/aion-3.0"},"aion-labs/aion-2.0":{"id":"aion-labs/aion-2.0","name":"AionLabs: Aion-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-12-21","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.8,"output":1.6,"cache_read":0.2},"canonical_model_id":"aion-labs/aion-2.0"},"aion-labs/aion-3.5":{"id":"aion-labs/aion-3.5","name":"AionLabs: Aion 3.5","description":"A GLM-family collaborative generation model tuned for immersive roleplay and storytelling, with stronger narrative structure, tension, conflict, and nuanced mature themes.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":3,"output":6,"cache_read":0.75}},"ateron/gemma-4-dark-thoughts-v2-31b":{"id":"ateron/gemma-4-dark-thoughts-v2-31b","name":"Gemma 4 Dark Thoughts V2 31B","description":"Ateron's second Dark Thoughts merge combines Dark Scarlett V2, MeroMero V2, and Scotoma 2 for creative dialogue and extended character roleplay. This text-only Gemma 4 31B deployment supports a 65,536-token context window and up to 8,192 output tokens.","family":"gemma","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.1,"output":0.45,"cache_read":0.05}},"huihui-ai/DeepSeek-R1-Distill-Llama-70B-abliterated":{"id":"huihui-ai/DeepSeek-R1-Distill-Llama-70B-abliterated","name":"DeepSeek R1 Llama 70B Abliterated","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-01-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"huihui-ai/Qwen2.5-32B-Instruct-abliterated":{"id":"huihui-ai/Qwen2.5-32B-Instruct-abliterated","name":"Qwen 2.5 32B Abliterated","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-01-06","last_updated":"2025-01-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"huihui-ai/Llama-3.3-70B-Instruct-abliterated":{"id":"huihui-ai/Llama-3.3-70B-Instruct-abliterated","name":"Llama 3.3 70B Instruct abliterated","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-08-08","last_updated":"2025-08-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated":{"id":"huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated","name":"DeepSeek R1 Qwen Abliterated","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-01-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":1.4,"output":1.4,"cache_read":0.7}},"nothingiisreal/L3.1-70B-Celeste-V0.1-BF16":{"id":"nothingiisreal/L3.1-70B-Celeste-V0.1-BF16","name":"Llama 3.1 70B Celeste v0.1","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5":{"id":"failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5","name":"Llama 3 70B abliterated","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":0.7,"output":0.7,"cache_read":0.35}},"inflatebot/MN-12B-Mag-Mell-R1":{"id":"inflatebot/MN-12B-Mag-Mell-R1","name":"Mag Mell R1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"moondream/moondream3.1-9b-a2b":{"id":"moondream/moondream3.1-9b-a2b","name":"Moondream 3.1 9B A2B","description":"Moondream 3.1 is the successor to Moondream 3 Preview: a distinct 9B mixture-of-experts vision-language checkpoint with 2B active parameters, for visual question answering, captioning, OCR, and visual reasoning. It accepts text or one image with optional reasoning, a 32,000-token context, and up to 4,096 output tokens.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"moondream/moondream3-preview":{"id":"moondream/moondream3-preview","name":"Moondream 3 Preview","description":"Moondream 3 Preview is the original 9B mixture-of-experts vision-language checkpoint with 2B active parameters, for visual question answering, captioning, OCR, and visual reasoning. It accepts text or one image with optional reasoning, a 32,000-token context, and up to 4,096 output tokens.","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":4096},"cost":{"input":0.3,"output":2.5}},"amazon/nova-2-lite-v1":{"id":"amazon/nova-2-lite-v1","name":"Amazon Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65535},"cost":{"input":0.51,"output":4.25,"cache_read":0.255},"canonical_model_id":"amazon/nova-2-lite-v1"},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Amazon Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"input":300000,"output":32000},"cost":{"input":0.799,"output":3.196,"cache_read":0.3995},"canonical_model_id":"amazon/nova-pro-v1"},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Amazon Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"input":300000,"output":5120},"cost":{"input":0.0595,"output":0.238,"cache_read":0.02975},"canonical_model_id":"amazon/nova-lite-v1"},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents","family":"ling","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.06,"output":0.18,"cache_read":0.012},"canonical_model_id":"inclusionai/ling-3.0-flash-vl"},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.0-flash:thinking":{"id":"inclusionai/ling-3.0-flash:thinking","name":"Ling 3.0 Flash Thinking","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.0-flash"},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.22,"cache_read":0.015},"canonical_model_id":"inclusionai/ling-3.1-flash"},"reka/reka-flash-3":{"id":"reka/reka-flash-3","name":"Reka Flash 3","description":"Reka's 21B reasoning model for chat and coding, served directly by Reka.","family":"reka","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":65536},"cost":{"input":0.105,"output":0.21},"canonical_model_id":"rekaai/reka-flash-3"},"reka/reka-edge-2603":{"id":"reka/reka-edge-2603","name":"Reka Edge 2603","description":"Reka's compact multimodal model for visual understanding and tool use, served directly by Reka.","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":16384},"cost":{"input":0.105,"output":0.105},"canonical_model_id":"rekaai/reka-edge"},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"input":6144,"output":4096},"cost":{"input":0.799,"output":1.207,"cache_read":0.3995},"canonical_model_id":"undi95/remm-slerp-l2-13b"},"deepseek-ai/DeepSeek-V3.1":{"id":"deepseek-ai/DeepSeek-V3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.2,"output":0.7,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/DeepSeek-V3.1:thinking":{"id":"deepseek-ai/DeepSeek-V3.1:thinking","name":"DeepSeek V3.1 Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.2,"output":0.7,"cache_read":0.1},"canonical_model_id":"deepseek/deepseek-v3.1"},"deepseek-ai/deepseek-v3.2-exp":{"id":"deepseek-ai/deepseek-v3.2-exp","name":"DeepSeek V3.2 Exp","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"input":163840,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.14}},"deepseek-ai/deepseek-v3.2-exp-thinking":{"id":"deepseek-ai/deepseek-v3.2-exp-thinking","name":"DeepSeek V3.2 Exp Thinking","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"input":163840,"output":65536},"cost":{"input":0.28,"output":0.42,"cache_read":0.14}},"deepseek-ai/DeepSeek-V3.1-Terminus":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus","name":"DeepSeek V3.1 Terminus","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2025-08-02","last_updated":"2025-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.25,"output":0.7,"cache_read":0.125}},"deepseek-ai/DeepSeek-R1-0528":{"id":"deepseek-ai/DeepSeek-R1-0528","name":"DeepSeek R1 0528","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"input":163840,"output":32768},"cost":{"input":0.5,"output":2.15,"cache_read":0.35},"canonical_model_id":"deepseek/deepseek-r1-0528"},"deepseek-ai/DeepSeek-V3.1-Terminus:thinking":{"id":"deepseek-ai/DeepSeek-V3.1-Terminus:thinking","name":"DeepSeek V3.1 Terminus (Thinking)","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek-thinking","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"release_date":"2024-01-01","last_updated":"2025-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.25,"output":0.7,"cache_read":0.125}},"Sao10K/L3.3-70B-Euryale-v2.3":{"id":"Sao10K/L3.3-70B-Euryale-v2.3","name":"Llama 3.3 70B Euryale","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-12-18","last_updated":"2024-12-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":20480,"input":20480,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465},"canonical_model_id":"sao10k/l3.3-euryale-70b"},"Sao10K/L3.1-70B-Hanami-x1":{"id":"Sao10K/L3.1-70B-Hanami-x1","name":"Llama 3.1 70B Hanami","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Sao10K/L3.1-70B-Euryale-v2.2":{"id":"Sao10K/L3.1-70B-Euryale-v2.2","name":"Llama 3.1 70B Euryale","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":20480,"input":20480,"output":16384},"cost":{"input":0.306,"output":0.357,"cache_read":0.153},"canonical_model_id":"sao10k/l3.1-euryale-70b"},"Sao10K/L3-8B-Stheno-v3.2":{"id":"Sao10K/L3-8B-Stheno-v3.2","name":"Sao10K Stheno 8b","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-11-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003}},"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1":{"id":"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1","name":"EVA-LLaMA-3.33-70B-v0.1","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":2.006,"output":2.006,"cache_read":1.003}},"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0":{"id":"EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0","name":"EVA Llama 3.33 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":16384},"cost":{"input":2.006,"output":2.006,"cache_read":1.003}},"EVA-UNIT-01/EVA-Qwen2.5-72B-v0.2":{"id":"EVA-UNIT-01/EVA-Qwen2.5-72B-v0.2","name":"EVA-Qwen2.5-72B-v0.2","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.799,"output":0.799,"cache_read":0.3995}},"EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2":{"id":"EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2","name":"EVA-Qwen2.5-32B-v0.2","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-07-26","last_updated":"2025-07-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.799,"output":0.799,"cache_read":0.3995}},"openai/gpt-6-astra-pro":{"id":"openai/gpt-6-astra-pro","name":"GPT 6 Astra Pro","description":"GPT 6 Astra in Pro reasoning mode. Uses additional model work for difficult tasks, with higher latency and token usage at the same per-token rates. Reasoning effort remains independently configurable.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/o1":{"id":"openai/o1","name":"OpenAI o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":15,"output":60,"cache_read":7.5},"canonical_model_id":"openai/o1"},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT 4.1 Mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1047576,"output":32768},"cost":{"input":0.4,"output":1.6,"cache_read":0.1},"canonical_model_id":"openai/gpt-4.1-mini"},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT 4.1 Nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1047576,"output":32768},"cost":{"input":0.1,"output":0.4,"cache_read":0.025},"canonical_model_id":"openai/gpt-4.1-nano"},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.03,"output":0.15,"cache_read":0.03},"canonical_model_id":"openai/gpt-oss-120b"},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT 5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"openai/o1-pro":{"id":"openai/o1-pro","name":"OpenAI o1 Pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":150,"output":600,"cache_read":75},"canonical_model_id":"openai/o1-pro"},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT 5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":2.5,"output":20,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.1-codex-max"},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT 5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2.5,"output":15,"cache_read":0.25},"canonical_model_id":"openai/gpt-5.4"},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT 6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"openai/gpt-6-astra"},"openai/gpt-5.6-luna-pro":{"id":"openai/gpt-5.6-luna-pro","name":"GPT 5.6 Luna Pro","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT 4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"input":1047576,"output":32768},"cost":{"input":2,"output":8,"cache_read":0.5},"canonical_model_id":"openai/gpt-4.1"},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT 5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1-codex"},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":16384},"cost":{"input":0.02,"output":0.1},"canonical_model_id":"openai/gpt-oss-20b"},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-08-06"},"openai/gpt-latest":{"id":"openai/gpt-latest","name":"GPT Latest","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT 5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5-mini"},"openai/o3-mini-low":{"id":"openai/o3-mini-low","name":"OpenAI o3-mini (Low)","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-sol-latest":{"id":"openai/gpt-sol-latest","name":"GPT Sol Latest","description":"Compatibility alias that routes to GPT 5.6 Sol, the latest supported GPT Sol model.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"openai/gpt-terra-latest":{"id":"openai/gpt-terra-latest","name":"GPT Terra Latest","description":"Compatibility alias that routes to GPT 5.6 Terra, the latest supported GPT Terra model.","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT 5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"openai/gpt-5.5"},"openai/gpt-6.1-sol-pro":{"id":"openai/gpt-6.1-sol-pro","name":"GPT 6.1 Sol Pro","description":"GPT-6.1 Sol Pro uses the same underlying model as GPT-6.1 Sol with Pro reasoning mode enabled for higher-quality responses on complex tasks.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT 6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125},"canonical_model_id":"openai/gpt-6-luna"},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o-2024-11-20"},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT 5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.05,"output":0.4,"cache_read":0.005},"canonical_model_id":"openai/gpt-5-nano"},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT 6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-6-sol"},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT 5.4 Mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.75,"output":4.5,"cache_read":0.075},"canonical_model_id":"openai/gpt-5.4-mini"},"openai/gpt-6-luna-pro":{"id":"openai/gpt-6-luna-pro","name":"GPT 6 Luna Pro","description":"GPT-6 Luna Pro uses the same underlying model as GPT-6 Luna with Pro reasoning mode enabled for higher-quality responses on complex tasks.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT 5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.3-codex"},"openai/o3-mini":{"id":"openai/o3-mini","name":"OpenAI o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/o4-mini-high":{"id":"openai/o4-mini-high","name":"OpenAI o4-mini high","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o4-mini"},"openai/gpt-chat-latest":{"id":"openai/gpt-chat-latest","name":"GPT Chat Latest","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT 5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2"},"openai/gpt-5.6-terra-pro":{"id":"openai/gpt-5.6-terra-pro","name":"GPT 5.6 Terra Pro","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-luna-latest":{"id":"openai/gpt-luna-latest","name":"GPT Luna Latest","description":"Compatibility alias that routes to GPT 5.6 Luna, the latest supported GPT Luna model.","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125}},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT 5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":12,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-terra"},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT 6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"openai/gpt-6.1-sol"},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT 5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":15,"output":120,"cache_read":1.5},"canonical_model_id":"openai/gpt-5-pro"},"openai/o4-mini":{"id":"openai/o4-mini","name":"OpenAI o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o4-mini"},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.075,"output":0.3},"canonical_model_id":"openai/gpt-oss-safeguard-20b"},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"input":16385,"output":4096},"cost":{"input":0.5,"output":1.5},"canonical_model_id":"openai/gpt-3.5-turbo"},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT 5.4 Nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.2,"output":1.25,"cache_read":0.02},"canonical_model_id":"openai/gpt-5.4-nano"},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT 5.1 Codex Mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":0.25,"output":2,"cache_read":0.025},"canonical_model_id":"openai/gpt-5.1-codex-mini"},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":10,"cache_read":1.25},"canonical_model_id":"openai/gpt-4o"},"openai/o3":{"id":"openai/o3","name":"OpenAI o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":2,"output":8,"cache_read":1},"canonical_model_id":"openai/o3"},"openai/gpt-5.6-sol-pro":{"id":"openai/gpt-5.6-sol-pro","name":"GPT 5.6 Sol Pro","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT 5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":0.2,"output":1.2,"cache_read":0.02,"cache_write":0.25},"canonical_model_id":"openai/gpt-5.6-luna"},"openai/gpt-6-sol-pro":{"id":"openai/gpt-6-sol-pro","name":"GPT 6 Sol Pro","description":"GPT-6 Sol Pro uses the same underlying model as GPT-6 Sol with Pro reasoning mode enabled for higher-quality responses on complex tasks.","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5}},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":0.15,"output":0.6,"cache_read":0.075},"canonical_model_id":"openai/gpt-4o-mini"},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT 5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.75,"output":14,"cache_read":0.175},"canonical_model_id":"openai/gpt-5.2-codex"},"openai/gpt-astra-latest":{"id":"openai/gpt-astra-latest","name":"GPT Astra Latest","description":"Compatibility alias that routes to GPT 6 Astra, the latest supported GPT Astra model.","family":"gpt-astra","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5}},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT 5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5"},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":10,"output":30},"canonical_model_id":"openai/gpt-4-turbo"},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT 5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":1050000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"openai/gpt-5.6-sol"},"openai/o3-pro-2025-06-10":{"id":"openai/o3-pro-2025-06-10","name":"OpenAI o3-pro (2025-06-10)","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":22,"output":88,"cache_read":11},"canonical_model_id":"openai/o3-pro"},"openai/o3-mini-high":{"id":"openai/o3-mini-high","name":"OpenAI o3-mini (High)","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-mini","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":100000},"cost":{"input":1.1,"output":4.4,"cache_read":0.55},"canonical_model_id":"openai/o3-mini"},"openai/gpt-5.1-2025-11-13":{"id":"openai/gpt-5.1-2025-11-13","name":"GPT-5.1 (2025-11-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":400000,"output":128000},"cost":{"input":1.25,"output":10,"cache_read":0.125},"canonical_model_id":"openai/gpt-5.1"},"stepfun-ai/step-3.5-flash-2603":{"id":"stepfun-ai/step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"stepfun/step-3.5-flash-2603"},"stepfun-ai/step-3.5-flash":{"id":"stepfun-ai/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"stepfun/step-3.5-flash"},"abliteration-ai/abliterated-model-large":{"id":"abliteration-ai/abliterated-model-large","name":"Abliterated Model Large","description":"Abliteration.ai's large text reasoning model is derived from GLM-5.2 and supports native tool calling, structured output, automatic prompt caching, and a one-million-token context window.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":3,"output":5,"cache_read":0.3}},"abliteration-ai/abliterated-model-large-v2":{"id":"abliteration-ai/abliterated-model-large-v2","name":"Abliterated Model Large V2","description":"Abliteration.ai's default large text reasoning model is derived from GLM-5.3 for harder reasoning and evaluation workloads, with automatic prompt caching and a one-million-token context window.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-08-31","last_updated":"2026-08-31","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":999990},"cost":{"input":3,"output":5,"cache_read":0.3}},"abliteration-ai/abliterated-model-large-flash":{"id":"abliteration-ai/abliterated-model-large-flash","name":"Abliterated Large Flash","description":"Abliterated Large Flash is an uncensored reasoning model based on DeepSeek V4.1-Flash-1017. It accepts text and images and returns text, with tool calling, JSON mode, JSON Schema output, and automatic prompt caching. Its shared input/output context is 1,048,576 tokens, with up to 262,144 output tokens when space permits. Reasoning defaults to high and can be turned off or set to low, high, or max. Supports up to four images across the request history.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":262144},"cost":{"input":1,"output":2,"cache_read":0.1}},"abliteration-ai/abliterated-model":{"id":"abliteration-ai/abliterated-model","name":"Abliterated Model","description":"Abliteration.ai's multimodal reasoning model supports text and image input, structured output, automatic prompt caching, and a 262K-token context window.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":false,"structured_output":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":262134},"cost":{"input":1,"output":3,"cache_read":0.1}},"prism-ml/ternary-bonsai-2-9b":{"id":"prism-ml/ternary-bonsai-2-9b","name":"Ternary Bonsai2 9B","description":"Ternary Bonsai2 9B is Prism ML’s open-weight model derived from Qwen3.5-9B with ternary language weights. It accepts text and generates text, supports reasoning and tool calling, and offers a shared 262,144-token context window with up to 65,536 output tokens. Thinking is enabled by default and can be turned off.","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.08,"output":0.13,"cache_read":0.04}},"prism-ml/ternary-bonsai-2-27b":{"id":"prism-ml/ternary-bonsai-2-27b","name":"Ternary Bonsai 2 27B","description":"Ternary Bonsai 2 27B is a 27B-parameter reasoning model from PrismML derived from Qwen3.8-27B. It supports coding, mathematics, tool calling, and image understanding with a 262K-token context window.","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["medium","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-09-18","last_updated":"2026-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.075,"output":0.5,"cache_read":0.0375}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.03,"output":0.15,"cache_read":0.015},"canonical_model_id":"inference-net/schematron-v2-turbo"},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":0.05,"output":0.23,"cache_read":0.025},"canonical_model_id":"inference-net/schematron-v2-small"},"microsoft/wizardlm-2-8x22b":{"id":"microsoft/wizardlm-2-8x22b","name":"WizardLM-2 8x22B","description":"Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-04-15","last_updated":"2024-04-15","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465},"canonical_model_id":"microsoft/wizardlm-2-8x22b"},"poolside/laguna-s-2.1:thinking":{"id":"poolside/laguna-s-2.1:thinking","name":"Laguna S 2.1 Thinking","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.01},"canonical_model_id":"poolside/laguna-s-2.1"},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":131072},"cost":{"input":0.1,"output":0.2,"cache_read":0.01},"canonical_model_id":"poolside/laguna-s-2.1"},"anthropic/claude-opus-4.1:thinking:1024":{"id":"anthropic/claude-opus-4.1:thinking:1024","name":"Claude 4.1 Opus Thinking (1K)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-sonnet-5:thinking":{"id":"anthropic/claude-sonnet-5:thinking","name":"Claude Sonnet 5 Thinking","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.7":{"id":"anthropic/claude-opus-4.7","name":"Claude 4.7 Opus","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-4.6:thinking:medium":{"id":"anthropic/claude-opus-4.6:thinking:medium","name":"Claude 4.6 Opus Thinking Medium","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-sonnet-4.5:thinking":{"id":"anthropic/claude-sonnet-4.5:thinking","name":"Claude Sonnet 4.5 Thinking","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-opus-4.1":{"id":"anthropic/claude-opus-4.1","name":"Claude 4.1 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.2,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5"},"anthropic/claude-opus-4.6:thinking:low":{"id":"anthropic/claude-opus-4.6:thinking:low","name":"Claude 4.6 Opus Thinking Low","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.8:thinking":{"id":"anthropic/claude-opus-4.8:thinking","name":"Claude Opus 4.8 Thinking","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-sonnet-5.5:thinking":{"id":"anthropic/claude-sonnet-5.5:thinking","name":"Claude Sonnet 5.5 Thinking","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-4.6:thinking":{"id":"anthropic/claude-opus-4.6:thinking","name":"Claude 4.6 Opus Thinking","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5,"cache_write":6.25},"canonical_model_id":"anthropic/claude-opus-5"},"anthropic/claude-haiku-4.5":{"id":"anthropic/claude-haiku-4.5","name":"Claude Haiku 4.5","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":1,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5"},"anthropic/claude-haiku-4.5:thinking":{"id":"anthropic/claude-haiku-4.5:thinking","name":"Claude Haiku 4.5 Thinking","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":1,"output":5,"cache_read":0.1},"canonical_model_id":"anthropic/claude-haiku-4-5"},"anthropic/claude-fable-latest":{"id":"anthropic/claude-fable-latest","name":"Claude Fable Latest","description":"Compatibility alias for Claude Fable.","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5}},"anthropic/claude-fable-5.1":{"id":"anthropic/claude-fable-5.1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":10,"output":50,"cache_read":0.25,"cache_write":12.5},"canonical_model_id":"anthropic/claude-fable-5-1"},"anthropic/claude-opus-4.7:thinking":{"id":"anthropic/claude-opus-4.7:thinking","name":"Claude 4.7 Opus Thinking","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-7"},"anthropic/claude-opus-5.5":{"id":"anthropic/claude-opus-5.5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5},"canonical_model_id":"anthropic/claude-opus-5-5"},"anthropic/claude-opus-4.1:thinking:8192":{"id":"anthropic/claude-opus-4.1:thinking:8192","name":"Claude 4.1 Opus Thinking (8K)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-opus-4.5:thinking":{"id":"anthropic/claude-opus-4.5:thinking","name":"Claude 4.5 Opus Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-latest":{"id":"anthropic/claude-sonnet-latest","name":"Claude Sonnet Latest","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-01","last_updated":"2026-03-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5}},"anthropic/claude-opus-4.5":{"id":"anthropic/claude-opus-4.5","name":"Claude 4.5 Opus","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-5"},"anthropic/claude-sonnet-4.5":{"id":"anthropic/claude-sonnet-4.5","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":64000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-5"},"anthropic/claude-sonnet-4.6:thinking":{"id":"anthropic/claude-sonnet-4.6:thinking","name":"Claude Sonnet 4.6 Thinking","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"anthropic/claude-opus-4.6":{"id":"anthropic/claude-opus-4.6","name":"Claude 4.6 Opus","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-opus-4.6:thinking:max":{"id":"anthropic/claude-opus-4.6:thinking:max","name":"Claude 4.6 Opus Thinking Max","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-6"},"anthropic/claude-haiku-5.5":{"id":"anthropic/claude-haiku-5.5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]},"canonical_model_id":"anthropic/claude-haiku-5-5"},"anthropic/claude-opus-4.1:thinking:32768":{"id":"anthropic/claude-opus-4.1:thinking:32768","name":"Claude 4.1 Opus Thinking (32K)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-haiku-latest":{"id":"anthropic/claude-haiku-latest","name":"Claude Haiku Latest","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":0.1,"output":0.5,"cache_read":0.01,"cache_write":0.125,"tiers":[{"input":0.5,"output":2.5,"cache_read":0.05,"cache_write":0.625,"tier":{"type":"context","size":100000}}]}},"anthropic/claude-sonnet-5.5":{"id":"anthropic/claude-sonnet-5.5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":10,"cache_read":0.1,"cache_write":2.5},"canonical_model_id":"anthropic/claude-sonnet-5-5"},"anthropic/claude-opus-latest":{"id":"anthropic/claude-opus-latest","name":"Claude Opus Latest","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-29","last_updated":"2026-03-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":4,"output":20,"cache_read":0.2,"cache_write":5}},"anthropic/claude-opus-4.8":{"id":"anthropic/claude-opus-4.8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":5,"output":25,"cache_read":0.5},"canonical_model_id":"anthropic/claude-opus-4-8"},"anthropic/claude-opus-4.1:thinking":{"id":"anthropic/claude-opus-4.1:thinking","name":"Claude 4.1 Opus Thinking","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"input":200000,"output":32000},"cost":{"input":15,"output":75,"cache_read":1.5},"canonical_model_id":"anthropic/claude-opus-4-1"},"anthropic/claude-sonnet-4.6":{"id":"anthropic/claude-sonnet-4.6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":3,"output":15,"cache_read":0.3},"canonical_model_id":"anthropic/claude-sonnet-4-6"},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":128000},"cost":{"input":2,"output":6,"cache_read":0.25},"canonical_model_id":"sakana/fugu-max"},"sakana/fugu-ultra-v1.1":{"id":"sakana/fugu-ultra-v1.1","name":"Fugu Ultra v1.1","description":"Sakana AI's upgraded Fugu Ultra release with stronger coding, agentic task execution, and advanced reasoning through dynamic orchestration of frontier models.","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":16384},"cost":{"input":5,"output":30,"cache_read":0.5}},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":16384},"cost":{"input":5,"output":30,"cache_read":0.5},"canonical_model_id":"sakana/fugu-ultra"},"minimax/minimax-01":{"id":"minimax/minimax-01","name":"MiniMax 01","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-01-15","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000192,"input":1000192,"output":16384},"cost":{"input":0.1394,"output":1.122,"cache_read":0.0697}},"minimax/minimax-m3.1-flash":{"id":"minimax/minimax-m3.1-flash","name":"MiniMax M3.1 Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":524288},"cost":{"input":0.15,"output":0.6,"cache_read":0.015},"canonical_model_id":"minimax/MiniMax-M3.1-Flash"},"minimax/minimax-m3:thinking":{"id":"minimax/minimax-m3:thinking","name":"MiniMax M3 Thinking","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"input":512000,"output":80000},"cost":{"input":0.23,"output":0.96,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-latest":{"id":"minimax/minimax-latest","name":"MiniMax Latest","description":"MiniMax multimodal coding model for long-context reasoning and agent tasks","family":"minimax","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"input":512000,"output":80000},"cost":{"input":0.23,"output":0.96,"cache_read":0.05}},"minimax/minimax-m3":{"id":"minimax/minimax-m3","name":"MiniMax M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":512000,"input":512000,"output":80000},"cost":{"input":0.23,"output":0.96,"cache_read":0.05},"canonical_model_id":"minimax/MiniMax-M3"},"minimax/minimax-m2.1":{"id":"minimax/minimax-m2.1","name":"MiniMax M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.33,"output":1.32,"cache_read":0.03},"canonical_model_id":"minimax/MiniMax-M2.1"},"minimax/minimax-m2-her":{"id":"minimax/minimax-m2-her","name":"MiniMax M2-her","description":"MiniMax M2 variant tuned for conversational and character-driven agent interactions","family":"minimax","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65532,"input":65532,"output":2048},"cost":{"input":0.302,"output":1.207,"cache_read":0.151},"canonical_model_id":"minimax/MiniMax-M2-Her"},"minimax/minimax-m2.7":{"id":"minimax/minimax-m2.7","name":"MiniMax M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":0.21,"output":0.84,"cache_read":0.042},"canonical_model_id":"minimax/MiniMax-M2.7"},"minimax/minimax-m2.5":{"id":"minimax/minimax-m2.5","name":"MiniMax M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":0.27,"output":1.08,"cache_read":0.027},"canonical_model_id":"minimax/MiniMax-M2.5"},"minimax/minimax-m2.7-turbo":{"id":"minimax/minimax-m2.7-turbo","name":"MiniMax M2.7 Turbo","description":"Efficient MiniMax model for quick assistance, coding, and routine automation","family":"minimax-m2.7","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"input":204800,"output":131072},"cost":{"input":0.6,"output":2.4,"cache_read":0.06}},"minimax/minimax-m2":{"id":"minimax/minimax-m2","name":"MiniMax M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"input":200000,"output":131072},"cost":{"input":0.17,"output":1.53,"cache_read":0.085},"canonical_model_id":"minimax/MiniMax-M2"},"Envoid/Llama-3.05-NT-Storybreaker-Ministral-70B":{"id":"Envoid/Llama-3.05-NT-Storybreaker-Ministral-70B","name":"Llama 3.05 Storybreaker Ministral 70b","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Envoid/Llama-3.05-Nemotron-Tenyxchat-Storybreaker-70B":{"id":"Envoid/Llama-3.05-Nemotron-Tenyxchat-Storybreaker-70B","name":"Nemotron Tenyxchat Storybreaker 70b","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"MarinaraSpaghetti/NemoMix-Unleashed-12B":{"id":"MarinaraSpaghetti/NemoMix-Unleashed-12B","name":"NemoMix 12B Unleashed","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":8192},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"Doctor-Shotgun/MS3.2-24B-Magnum-Diamond":{"id":"Doctor-Shotgun/MS3.2-24B-Magnum-Diamond","name":"MS3.2 24B Magnum Diamond","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.493,"output":0.493,"cache_read":0.2465}},"qwen/qwen3.7-max:thinking":{"id":"qwen/qwen3.7-max:thinking","name":"Qwen3.7 Max Thinking","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3.8-27b-novelist":{"id":"qwen/qwen3.8-27b-novelist","name":"Qwen 3.8 27B Novelist","description":"Qwen 3.8 27B Novelist is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen-turbo":{"id":"qwen/qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":8192},"cost":{"input":0.04998,"output":0.2006,"cache_read":0.02499},"canonical_model_id":"alibaba/qwen-turbo"},"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.05,"output":0.15,"cache_read":0.025},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3-vl-235b-a22b-instruct":{"id":"qwen/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":32768},"cost":{"input":0.3,"output":1.9,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-instruct"},"qwen/qwen3.8-27b-synthia":{"id":"qwen/qwen3.8-27b-synthia","name":"Qwen 3.8 27B Synthia","description":"Qwen 3.8 27B Synthia is a text-only roleplay finetune served by Aoru in NVFP4 with a 262,144-token context window.","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen-long":{"id":"qwen/qwen-long","name":"Qwen Long 10M","description":"Alibaba's huge context window model. Takes in up to 10 million tokens, which is equivalent to dozens of books.","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":10000000,"input":10000000,"output":8192},"cost":{"input":0.1003,"output":0.408,"cache_read":0.05015}},"qwen/qwen3.7-plus":{"id":"qwen/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.8-27b-queen":{"id":"qwen/qwen3.8-27b-queen","name":"Qwen 3.8 27B Queen","description":"Qwen 3.8 27B Queen is an open-weight roleplay finetune with image understanding, tool calling, optional reasoning, and a 262,144-token context window.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.5,"cache_read":0.125}},"qwen/qwen3-coder":{"id":"qwen/qwen3-coder","name":"Qwen 3 Coder 480B","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262000,"input":262000,"output":65536},"cost":{"input":0.13,"output":0.5,"cache_read":0.065},"canonical_model_id":"alibaba/qwen3-coder-480b-a35b-instruct"},"qwen/qwen-plus":{"id":"qwen/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":995904,"input":995904,"output":32768},"cost":{"input":0.3995,"output":1.2002,"cache_read":0.19975},"canonical_model_id":"alibaba/qwen-plus"},"qwen/qwen3-coder-30b-a3b-instruct":{"id":"qwen/qwen3-coder-30b-a3b-instruct","name":"Qwen3 Coder 30B A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-coder-30b-a3b-instruct"},"qwen/qwen3.8-max-prime":{"id":"qwen/qwen3.8-max-prime","name":"Qwen3.8 Max Prime","description":"High-throughput edition of Qwen3.8 Max for coding, professional work, multimodal understanding, and long-running agent workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["minimal","low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":131072},"cost":{"input":4,"output":12,"cache_read":0.5},"canonical_model_id":"alibaba/qwen3.8-max-prime"},"qwen/qwen3.5-omni-flash":{"id":"qwen/qwen3.5-omni-flash","name":"Qwen3.5 Omni Flash","description":"Qwen3.5 Omni Flash is Qwen's fast multimodal model. We verified live support for text prompts, images, audio files, and direct video URLs on Alibaba's chat-completions-compatible API. Alibaba describes Flash as a fully evolved version of Qwen3 Omni with audio input support across 60+ languages.","family":"qwen3.5","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-03-30","last_updated":"2026-03-30","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":49152,"input":49152,"output":16384}},"qwen/qwen-max":{"id":"qwen/qwen-max","name":"Qwen 2.5 Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":1.5997,"output":6.392,"cache_read":0.79985},"canonical_model_id":"alibaba/qwen-max"},"qwen/qwen-2.5-72b-instruct":{"id":"qwen/qwen-2.5-72b-instruct","name":"Qwen2.5 72B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2025-07-03","last_updated":"2025-07-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":8192},"cost":{"input":0.36,"output":0.4,"cache_read":0.1785}},"qwen/qwen3.5-flash:thinking":{"id":"qwen/qwen3.5-flash:thinking","name":"Qwen3.5 Flash Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.8-27b-fable":{"id":"qwen/qwen3.8-27b-fable","name":"Qwen 3.8 27B Fable","description":"Qwen 3.8 27B Fable is an open-weight multimodal creative finetune for expressive dialogue, long-form storytelling, character work, and roleplay.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-07-29","last_updated":"2026-08-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.5,"cache_read":0.125}},"qwen/qwen3.8-27b-obliterated:thinking":{"id":"qwen/qwen3.8-27b-obliterated:thinking","name":"Qwen 3.8 27B Obliterated Thinking","description":"Qwen 3.8 27B Obliterated with thinking enabled for more deliberate creative work, coding, multimodal analysis, tool use, and long-context problem solving.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.75,"cache_read":0.225}},"qwen/qwen3.6-27b":{"id":"qwen/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.203,"output":2.24,"cache_read":0.1015},"canonical_model_id":"alibaba/qwen3.6-27b"},"qwen/qwen3-30b-a3b-instruct-2507":{"id":"qwen/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B A3B Instruct 2507","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.2,"output":0.5,"cache_read":0.1},"canonical_model_id":"alibaba/qwen3-30b-a3b-instruct-2507"},"qwen/qwen3.7-flash:thinking":{"id":"qwen/qwen3.7-flash:thinking","name":"Qwen3.7 Flash Thinking","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3-coder-next":{"id":"qwen/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.250251,"output":0.80293,"cache_read":0.079764},"canonical_model_id":"alibaba/qwen3-coder-next"},"qwen/qwen3.8-flash":{"id":"qwen/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":131072},"cost":{"input":0.14,"output":0.42,"cache_read":0.016,"cache_write":0.2},"canonical_model_id":"alibaba/qwen3.8-flash"},"qwen/qwen3.8-omni-flash":{"id":"qwen/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":131072},"cost":{"input":0.15,"output":0.47,"cache_read":0.016},"canonical_model_id":"alibaba/qwen3.8-omni-flash"},"qwen/qwen2.5-coder-32b-instruct":{"id":"qwen/qwen2.5-coder-32b-instruct","name":"Qwen 2.5 Coder 32b","description":"Open coding-focused Qwen model for code generation, repair, and repository reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8192},"cost":{"input":0.2006,"output":0.2006,"cache_read":0.1003},"canonical_model_id":"alibaba/qwen2.5-coder-32b-instruct"},"qwen/qwen3.6-plus":{"id":"qwen/qwen3.6-plus","name":"Qwen 3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.325,"output":1.95,"cache_read":0.0325,"cache_write":0.40625},"canonical_model_id":"alibaba/qwen3.6-plus"},"qwen/qwen3-next-80b-a3b-instruct":{"id":"qwen/qwen3-next-80b-a3b-instruct","name":"Qwen3 Next 80B A3B (Instruct)","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":1.1,"cache_read":0.07},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-instruct"},"qwen/qwen3.5-122b-a10b:thinking":{"id":"qwen/qwen3.5-122b-a10b:thinking","name":"Qwen3.5 122B A10B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.26,"output":2.08,"cache_read":0.103788},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-235b-a22b-instruct-2507":{"id":"qwen/qwen3-235b-a22b-instruct-2507","name":"Qwen 3 235b A22B 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.0875,"output":0.35,"cache_read":0.0175},"canonical_model_id":"alibaba/qwen3-235b-a22b-instruct-2507"},"qwen/qwen3.5-397b-a17b:thinking":{"id":"qwen/qwen3.5-397b-a17b:thinking","name":"Qwen3.5 397B A17B Thinking","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.39,"output":2.34,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.8-27b-kiwen":{"id":"qwen/qwen3.8-27b-kiwen","name":"Qwen 3.8 27B Kiwen","description":"Qwen 3.8 27B Kiwen is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.5-flash":{"id":"qwen/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.1,"output":0.4,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-flash"},"qwen/qwen3.5-plus":{"id":"qwen/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.4,"output":2.4,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.39,"output":2.34,"cache_read":0.3},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.26,"output":2.08,"cache_read":0.103788},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3-14b":{"id":"qwen/qwen3-14b","name":"Qwen 3 14b","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-01-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.08,"output":0.24,"cache_read":0.04}},"qwen/qwen3.7-plus:thinking":{"id":"qwen/qwen3.7-plus:thinking","name":"Qwen3.7 Plus Thinking","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.4,"output":1.6,"cache_read":0.08,"cache_write":0.5},"canonical_model_id":"alibaba/qwen3.7-plus"},"qwen/qwen3.8-max:thinking":{"id":"qwen/qwen3.8-max:thinking","name":"Qwen3.8 Max Thinking","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"input":991000,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-27b-kimiko":{"id":"qwen/qwen3.8-27b-kimiko","name":"Qwen 3.8 27B Kimiko","description":"Qwen 3.8 27B Kimiko is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.8-27b-hemmingway":{"id":"qwen/qwen3.8-27b-hemmingway","name":"Qwen 3.8 27B Hemingway","description":"Qwen 3.8 27B Hemingway is an open-weight NVFP4 multimodal creative finetune for long-form prose, character dialogue, storytelling, and roleplay.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.25,"output":1.5,"cache_read":0.125}},"qwen/qwen3.6-flash":{"id":"qwen/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.19,"output":1.16,"cache_read":0.02,"cache_write":0.24},"canonical_model_id":"alibaba/qwen3.6-flash"},"qwen/qwen3.8-27b-fimi":{"id":"qwen/qwen3.8-27b-fimi","name":"Qwen 3.8 27B Fimi","description":"Qwen 3.8 27B Fimi is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3-max":{"id":"qwen/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.2002,"output":6.001,"cache_read":0.6001},"canonical_model_id":"alibaba/qwen3-max"},"qwen/qwen3.8-27b-antislop":{"id":"qwen/qwen3.8-27b-antislop","name":"Qwen 3.8 27B Antislop","description":"Qwen 3.8 27B Antislop is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.8-max-0902":{"id":"qwen/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":131072},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max-0902"},"qwen/qwen3.5-35b-a3b:thinking":{"id":"qwen/qwen3.5-35b-a3b:thinking","name":"Qwen3.5 35B A3B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.8-27b-uncensored:thinking":{"id":"qwen/qwen3.8-27b-uncensored:thinking","name":"Qwen 3.8 27B Uncensored Thinking","description":"Qwen 3.8 27B Uncensored with thinking enabled for more deliberate creative work, coding, multimodal analysis, tool use, and long-context problem solving.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1,"output":3,"cache_read":0.22}},"qwen/qwen3-32b":{"id":"qwen/qwen3-32b","name":"Qwen 3 32b","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-32b"},"qwen/qwen3.7-max":{"id":"qwen/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":65536},"cost":{"input":1.475,"output":4.425,"cache_read":0.295,"cache_write":1.84375},"canonical_model_id":"alibaba/qwen3.7-max"},"qwen/qwen3-next-80b-a3b-thinking":{"id":"qwen/qwen3-next-80b-a3b-thinking","name":"Qwen3 Next 80B A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":0.15,"output":0.65,"cache_read":0.075},"canonical_model_id":"alibaba/qwen3-next-80b-a3b-thinking"},"qwen/qwen3.5-plus:thinking":{"id":"qwen/qwen3.5-plus:thinking","name":"Qwen3.5 Plus Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536},"cost":{"input":0.4,"output":2.4,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.5-plus"},"qwen/qwen3-235b-a22b-thinking-2507":{"id":"qwen/qwen3-235b-a22b-thinking-2507","name":"Qwen 3 235b A22B 2507 Thinking","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"input":131072,"output":81920},"cost":{"input":0.3,"output":0.5},"canonical_model_id":"alibaba/qwen3-235b-a22b-thinking-2507"},"qwen/qwen3.8-27b-uncensored":{"id":"qwen/qwen3.8-27b-uncensored","name":"Qwen 3.8 27B Uncensored","description":"Qwen 3.8 27B Uncensored is an NVFP4 open-weight multimodal model LoRA-tuned for fewer refusals across chat, coding, reasoning, tool use, and long-context work.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1,"output":3,"cache_read":0.22}},"qwen/qwen3-235b-a22b":{"id":"qwen/qwen3-235b-a22b","name":"Qwen 3 235b A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":16384},"cost":{"input":0.0875,"output":0.35,"cache_read":0.0175},"canonical_model_id":"alibaba/qwen3-235b-a22b"},"qwen/qwen3.5-27b:thinking":{"id":"qwen/qwen3.5-27b:thinking","name":"Qwen3.5 27B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.135},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.6-35b-a3b:thinking":{"id":"qwen/qwen3.6-35b-a3b:thinking","name":"Qwen3.6 35B A3B Thinking","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":16384},"cost":{"input":0.112,"output":0.8,"cache_read":0.056},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.6-35b-a3b":{"id":"qwen/qwen3.6-35b-a3b","name":"Qwen3.6 35B A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":16384},"cost":{"input":0.112,"output":0.8,"cache_read":0.056},"canonical_model_id":"alibaba/qwen3.6-35b-a3b"},"qwen/qwen3.8-max":{"id":"qwen/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":991000,"input":991000,"output":65536},"cost":{"input":2,"output":6,"cache_read":0.25,"cache_write":2.5},"canonical_model_id":"alibaba/qwen3.8-max"},"qwen/qwen3.8-27b-swift":{"id":"qwen/qwen3.8-27b-swift","name":"Qwen 3.8 27B Swift","description":"Qwen 3.8 27B Swift is a finetune for fast responses, served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.8-27b:thinking":{"id":"qwen/qwen3.8-27b:thinking","name":"Qwen3.8 27B Thinking","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.7,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.5-omni-plus":{"id":"qwen/qwen3.5-omni-plus","name":"Qwen3.5 Omni Plus","description":"Qwen3.5 Omni Plus is Qwen's stronger general multimodal model. We verified live support for text prompts, images, audio files, and direct video URLs on Alibaba's chat-completions-compatible API. Alibaba describes Plus as a comprehensive evolution of Qwen3 Omni with support for over 10 hours of audio input.","family":"qwen3.5","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-03-30","last_updated":"2026-03-30","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":983616,"input":983616,"output":65536}},"qwen/qwen3-30b-a3b":{"id":"qwen/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.1,"output":0.3,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3-30b-a3b"},"qwen/qwen3-max-2026-01-23":{"id":"qwen/qwen3-max-2026-01-23","name":"Qwen3 Max 2026-01-23","description":"Qwen3 Max is Alibaba's flagship Qwen 3 reasoning model with native tool use (web search, web extractor, code interpreter) and a 256K context window.","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2026-01-26","last_updated":"2026-01-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":32768},"cost":{"input":1.2002,"output":6.001,"cache_read":0.6001}},"qwen/qwen3.6-max-preview":{"id":"qwen/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":245760,"input":245760,"output":65536},"cost":{"input":1.04,"output":6.24,"cache_read":0.52},"canonical_model_id":"alibaba/qwen3.6-max-preview"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.14,"output":1,"cache_read":0.05},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3-vl-235b-a22b-thinking":{"id":"qwen/qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.5,"output":6,"cache_read":0.25},"canonical_model_id":"alibaba/qwen3-vl-235b-a22b-thinking"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":260096,"input":260096,"output":65536},"cost":{"input":0.25,"output":2,"cache_read":0.135},"canonical_model_id":"alibaba/qwen3.5-27b"},"qwen/qwen3.8-27b-obliterated":{"id":"qwen/qwen3.8-27b-obliterated","name":"Qwen 3.8 27B Obliterated","description":"Qwen 3.8 27B Obliterated is an open-weight multimodal model LoRA-tuned for fewer refusals across chat, coding, reasoning, tool use, and long-context work.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"release_date":"2026-07-29","last_updated":"2026-08-24","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"input":524288,"output":32768},"cost":{"input":0.25,"output":1.75,"cache_read":0.225}},"qwen/qwen3-8b":{"id":"qwen/qwen3-8b","name":"Qwen 3 8B","description":"Dense 8B Qwen3 model with switchable thinking for chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":41000,"input":41000,"output":32768},"cost":{"input":0.47,"output":0.47,"cache_read":0.235},"canonical_model_id":"alibaba/qwen3-8b"},"qwen/qwen3.8-27b-cybersecurity":{"id":"qwen/qwen3.8-27b-cybersecurity","name":"Qwen 3.8 27B Cybersecurity","description":"Qwen 3.8 27B Cybersecurity is a cybersecurity-focused variant based on the uncensored model, with provider moderation for illegal activities. It supports optional reasoning, image understanding, tool calling, and a 262,144-token context window.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-09-19","last_updated":"2026-09-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3.8-27b-darkidol":{"id":"qwen/qwen3.8-27b-darkidol","name":"Qwen 3.8 27B DarkIdol","description":"Qwen 3.8 27B DarkIdol is a creative finetune served by Aoru in NVFP4 with a 262,144-token context window. It supports image understanding, tool calling, and optional reasoning.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"release_date":"2026-10-05","last_updated":"2026-10-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":32768},"cost":{"input":0.1,"output":0.6,"cache_read":0.05}},"qwen/qwen3-coder-flash":{"id":"qwen/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.3,"output":1.5,"cache_read":0.15},"canonical_model_id":"alibaba/qwen3-coder-flash"},"qwen/qwen3.8-27b":{"id":"qwen/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"input":1048576,"output":32768},"cost":{"input":0.15,"output":0.7,"cache_read":0.04},"canonical_model_id":"alibaba/qwen3.8-27b"},"qwen/qwen3.7-flash":{"id":"qwen/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":991808,"input":991808,"output":65536},"cost":{"input":0.03,"output":0.13,"cache_read":0.006,"cache_write":0.038},"canonical_model_id":"alibaba/qwen3.7-flash"},"qwen/qwen3.6-27b:thinking":{"id":"qwen/qwen3.6-27b:thinking","name":"Qwen3.6 27B Thinking","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":false,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"input":262144,"output":65536},"cost":{"input":0.203,"output":2.24,"cache_read":0.1015},"canonical_model_id":"alibaba/qwen3.6-27b"},"meganova-ai/manta-flash-1.0":{"id":"meganova-ai/manta-flash-1.0","name":"Manta Flash 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-20","last_updated":"2025-12-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"input":16384,"output":16384},"cost":{"input":0.02,"output":0.16,"cache_read":0.01}},"meganova-ai/manta-mini-1.0":{"id":"meganova-ai/manta-mini-1.0","name":"Manta Mini 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-20","last_updated":"2025-12-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"input":8192,"output":8192},"cost":{"input":0.02,"output":0.16,"cache_read":0.01}},"meganova-ai/manta-pro-1.0":{"id":"meganova-ai/manta-pro-1.0","name":"Manta Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-09-20","last_updated":"2025-12-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"input":65536,"output":32768},"cost":{"input":0.06,"output":0.5,"cache_read":0.03}},"THUDM/GLM-4-9B-0414":{"id":"THUDM/GLM-4-9B-0414","name":"GLM 4 9B 0414","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8000},"cost":{"input":0.2,"output":0.2,"cache_read":0.1}},"THUDM/GLM-Z1-9B-0414":{"id":"THUDM/GLM-Z1-9B-0414","name":"GLM Z1 9B 0414","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm-z","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"input":32000,"output":8000},"cost":{"input":0.2,"output":0.2,"cache_read":0.1}},"THUDM/GLM-4-32B-0414":{"id":"THUDM/GLM-4-32B-0414","name":"GLM 4 32B 0414","description":"Flagship GLM model for hybrid reasoning, coding, and agentic engineering","family":"glm","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2025-04-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":65536},"cost":{"input":0.2,"output":0.2,"cache_read":0.1}},"ReadyArt/MS3.2-The-Omega-Directive-24B-Unslop-v2.0":{"id":"ReadyArt/MS3.2-The-Omega-Directive-24B-Unslop-v2.0","name":"Omega Directive 24B Unslop v2.0","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"input":32768,"output":32768},"cost":{"input":0.5,"output":0.5,"cache_read":0.25}},"x-ai/grok-4.7":{"id":"x-ai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":500000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.7"},"x-ai/grok-4.6":{"id":"x-ai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5},"canonical_model_id":"xai/grok-4.6"},"x-ai/grok-4.20-multi-agent":{"id":"x-ai/grok-4.20-multi-agent","name":"Grok 4.20 Multi-Agent","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"input":2000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-4.3":{"id":"x-ai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":1000000,"output":900000},"cost":{"input":1.25,"output":2.5,"cache_read":0.2},"canonical_model_id":"xai/grok-4.3"},"x-ai/grok-build-0.1":{"id":"x-ai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":230400},"cost":{"input":1,"output":2,"cache_read":0.2},"canonical_model_id":"xai/grok-build-0.1"},"x-ai/grok-4.20":{"id":"x-ai/grok-4.20","name":"Grok 4.20","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"input":2000000,"output":131072},"cost":{"input":1.25,"output":2.5,"cache_read":0.2}},"x-ai/grok-latest":{"id":"x-ai/grok-latest","name":"Grok Latest","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high","xhigh"]}],"tool_call":true,"structured_output":true,"release_date":"2026-05-03","last_updated":"2026-05-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.5}},"x-ai/grok-4.5":{"id":"x-ai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"input":500000,"output":450000},"cost":{"input":2,"output":6,"cache_read":0.3},"canonical_model_id":"xai/grok-4.5"},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"ByteDance Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.25,"output":2,"cache_read":0.125},"canonical_model_id":"bytedance-seed/seed-2.0-lite"},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"ByteDance Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":131072},"cost":{"input":0.5,"output":3,"cache_read":0.25},"canonical_model_id":"bytedance-seed/seed-2.0-code"},"bytedance-seed/seed-2-1-turbo":{"id":"bytedance-seed/seed-2-1-turbo","name":"ByteDance Seed 2.1 Turbo","description":"ByteDance Seed 2.1 Turbo is a multimodal model for coding and long-horizon agent workflows, including end-to-end software delivery and multi-step task execution. It supports text, image, and video input with a 262k context window.","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"input":262144,"output":235929},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":true,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":2.006,"output":2.992,"cache_read":1.003},"canonical_model_id":"anthracite-org/magnum-v4-72b"},"anthracite-org/magnum-v2-72b":{"id":"anthracite-org/magnum-v2-72b","name":"Magnum V2 72B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2024-01-01","last_updated":"2024-07-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"input":16384,"output":8192},"cost":{"input":2.006,"output":2.992,"cache_read":1.003}},"bytedance/doubao-seed-2.1-turbo":{"id":"bytedance/doubao-seed-2.1-turbo","name":"Doubao Seed 2.1 Turbo","description":"Fast, lower-cost model in the Doubao Seed 2.1 family for everyday chat, coding assistance, document work, and high-throughput productivity tasks. Supports a 256k context window and up to 128k output tokens. Note: privacy and logging guarantees may be limited.","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":0.5,"output":2.5,"cache_read":0.25},"canonical_model_id":"bytedance-seed/seed-2.1-turbo"},"bytedance/doubao-seed-2.1-pro":{"id":"bytedance/doubao-seed-2.1-pro","name":"Doubao Seed 2.1 Pro","description":"Higher-capability model in the Doubao Seed 2.1 family for agentic coding, long-context analysis, complex instruction following, and productivity workflows. Supports a 256k context window and up to 128k output tokens. Note: privacy and logging guarantees may be limited.","family":"seed","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"input":256000,"output":128000},"cost":{"input":1,"output":5,"cache_read":0.5},"canonical_model_id":"bytedance-seed/seed-2.1-pro"},"bytedance/doubao-seed-character":{"id":"bytedance/doubao-seed-character","name":"Doubao Seed Character","description":"ByteDance's character-focused Doubao Seed model for roleplay, persona consistency, dialogue, and creative character interactions. It supports text and image input with a 128k context window. Requests route through ZenMux to ByteDance; ZenMux does not publish a model-API zero-retention or training guarantee, so avoid sensitive data.","family":"seed","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":false,"structured_output":false,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":128000,"output":32768},"cost":{"input":0.15,"output":0.37,"cache_read":0.03,"cache_write":0.003},"canonical_model_id":"bytedance-seed/seed-character"},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Cohere: Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":4096},"cost":{"input":2.856,"output":14.246,"cache_read":1.428},"canonical_model_id":"cohere/command-r-plus-08-2024"},"heabsy/qwen3.8-27b-uncensored-fast":{"id":"heabsy/qwen3.8-27b-uncensored-fast","name":"Qwen 3.8 27B Uncensored Fast by Heabsy","description":"Heabsy's NVFP4 build of Qwen3.8-27B by Alibaba Cloud, with refusal removal based on huihui-ai's Huihui-Qwen3.8-27B-abliterated. Quantized and calibrated by Heabsy, with unpublished weights under the Heabsy Model License. Supports text and image input, tools, and optional reasoning capped at 16,384 tokens within the output budget. This named build is separate from Qwen 3.8 27B Uncensored.","family":"qwen","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-10-16","last_updated":"2026-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":235929},"cost":{"input":1,"output":3,"cache_read":0.22}},"heabsy/cyberheabsy":{"id":"heabsy/cyberheabsy","name":"CyberHeabsy","description":"A low-refusal text model for authorized security research, roleplay, and creative writing. It preserves tool calling after refusal removal, supports optional reasoning, and has a 1M-token context window.","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","low","medium","high"]}],"tool_call":true,"structured_output":false,"release_date":"2026-09-27","last_updated":"2026-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"input":1048576,"output":65536},"cost":{"input":1.5,"output":4.5,"cache_read":0.275}},"NousResearch/hermes-3-llama-3.1-70b":{"id":"NousResearch/hermes-3-llama-3.1-70b","name":"Hermes 3 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":true,"structured_output":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"input":65536,"output":8192},"cost":{"input":0.408,"output":0.408,"cache_read":0.204},"canonical_model_id":"nousresearch/hermes-3-llama-3.1-70b"},"NousResearch/hermes-4-405b:thinking":{"id":"NousResearch/hermes-4-405b:thinking","name":"Hermes 4 Large (Thinking)","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.15},"canonical_model_id":"nousresearch/hermes-4-405b"},"NousResearch/hermes-4-405b":{"id":"NousResearch/hermes-4-405b","name":"Hermes 4 Large","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":false,"tool_call":false,"structured_output":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":8192},"cost":{"input":0.3,"output":1.2,"cache_read":0.15},"canonical_model_id":"nousresearch/hermes-4-405b"},"Salesforce/Llama-xLAM-2-70b-fc-r":{"id":"Salesforce/Llama-xLAM-2-70b-fc-r","name":"Llama-xLAM-2 70B fc-r","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"release_date":"2025-04-13","last_updated":"2025-04-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"input":128000,"output":16384},"cost":{"input":2.5,"output":2.5,"cache_read":1.25}}}},"inceptron":{"id":"inceptron","env":["INCEPTRON_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.inceptron.io/v1","name":"Inceptron","doc":"https://docs.inceptron.io","models":{"moonshotai/Kimi-K2.6":{"id":"moonshotai/Kimi-K2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6026,"output":3.2,"cache_read":0.169,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.6"},"moonshotai/Kimi-K2.7-Code":{"id":"moonshotai/Kimi-K2.7-Code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"reasoning_options":[],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6712,"output":3.35,"cache_read":0.15,"cache_write":0},"canonical_model_id":"moonshotai/kimi-k2.7-code"},"deepseek-ai/DeepSeek-V4-Flash-0731":{"id":"deepseek-ai/DeepSeek-V4-Flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.11,"output":0.3666,"cache_read":0.027,"cache_write":0},"canonical_model_id":"deepseek/deepseek-v4-flash-0731"},"zai-org/GLM-5.3":{"id":"zai-org/GLM-5.3","name":"GLM 5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.6,"output":2.8,"cache_read":0.12,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3"},"zai-org/GLM-5.3-Flash":{"id":"zai-org/GLM-5.3-Flash","name":"GLM 5.3 Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.2,"output":0.75,"cache_read":0.075,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.3-flash"},"zai-org/GLM-5.2":{"id":"zai-org/GLM-5.2","name":"GLM 5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["high","max"]}],"tool_call":true,"interleaved":{"field":"reasoning_content"},"structured_output":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"cost":{"input":0.99,"output":1.188,"cache_read":0.16,"cache_write":0},"canonical_model_id":"zhipuai/glm-5.2"}}},"morph":{"id":"morph","env":["MORPH_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.morphllm.com/v1","name":"Morph","doc":"https://docs.morphllm.com/api-reference/introduction","models":{"morph-v3-large":{"id":"morph-v3-large","name":"Morph v3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.9,"output":1.9}},"morph-v3-fast":{"id":"morph-v3-fast","name":"Morph v3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-15","last_updated":"2024-08-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":16000},"cost":{"input":0.8,"output":1.2}},"auto":{"id":"auto","name":"Auto","description":"Automatic model router for matching prompts to suitable backends and budgets","family":"auto","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-06-01","last_updated":"2024-06-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":32000},"cost":{"input":0.85,"output":1.55}}}},"tencent-token-plan":{"id":"tencent-token-plan","env":["TENCENT_TOKEN_PLAN_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://api.lkeap.cloud.tencent.com/plan/v3","name":"Tencent Token Plan","doc":"https://cloud.tencent.com/document/product/1823/130060","models":{"hy4-preview":{"id":"hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000},"cost":{"input":0.834,"output":2.501,"cache_read":0.042},"canonical_model_id":"tencent/hy4-preview"},"hy3":{"id":"hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"},{"type":"effort","values":["none","high"]}],"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"cost":{"input":0,"output":0,"cache_read":0,"cache_write":0},"canonical_model_id":"tencent/hy3"}}},"tinfoil":{"id":"tinfoil","env":["TINFOIL_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://inference.tinfoil.sh/v1","name":"Tinfoil","doc":"https://docs.tinfoil.sh","models":{"gpt-oss-120b":{"id":"gpt-oss-120b","name":"gpt-oss-120b","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","medium","high"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-120b"},"gpt-oss-safeguard-120b":{"id":"gpt-oss-safeguard-120b","name":"gpt-oss-safeguard-120b","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":false,"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"cost":{"input":0.15,"output":0.6},"canonical_model_id":"openai/gpt-oss-safeguard-120b"},"deepseek-v4-1-flash":{"id":"deepseek-v4-1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":384000},"status":"beta","cost":{"input":0.65,"output":1.45,"cache_read":0.13},"canonical_model_id":"deepseek/deepseek-v4.1-flash"},"nomic-embed-text":{"id":"nomic-embed-text","name":"Nomic Embed Text v1.5","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","attachment":false,"reasoning":false,"tool_call":false,"structured_output":false,"temperature":false,"release_date":"2024-02","last_updated":"2024-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":768},"cost":{"input":0.05,"output":0}},"kimi-k3":{"id":"kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"cost":{"input":4,"output":20,"cache_read":0.8},"canonical_model_id":"moonshotai/kimi-k3"},"gemma4-31b":{"id":"gemma4-31b","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"reasoning_options":[{"type":"effort","values":["none","minimal","low","medium","high","xhigh","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"cost":{"input":0.4,"output":1},"canonical_model_id":"google/gemma-4-31b-it"},"llama3-3-70b":{"id":"llama3-3-70b","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":4096},"cost":{"input":1.75,"output":2.75},"canonical_model_id":"meta/llama-3.3-70b-instruct"},"glm-5-3":{"id":"glm-5-3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"reasoning_options":[{"type":"effort","values":["low","high","max"]}],"tool_call":true,"structured_output":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"cost":{"input":1.8,"output":5.75,"cache_read":0.45},"canonical_model_id":"zhipuai/glm-5.3"}}},"mixlayer":{"id":"mixlayer","env":["MIXLAYER_API_KEY"],"npm":"@ai-sdk/openai-compatible","api":"https://models.mixlayer.ai/v1","name":"Mixlayer","doc":"https://docs.mixlayer.com","models":{"qwen/qwen3.5-9b":{"id":"qwen/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.1,"output":0.4},"canonical_model_id":"alibaba/qwen3.5-9b"},"qwen/qwen3.5-397b-a17b":{"id":"qwen/qwen3.5-397b-a17b","name":"Qwen3.5 397B A17B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.6,"output":3.6},"canonical_model_id":"alibaba/qwen3.5-397b-a17b"},"qwen/qwen3.5-122b-a10b":{"id":"qwen/qwen3.5-122b-a10b","name":"Qwen3.5 122B A10B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.4,"output":3.2},"canonical_model_id":"alibaba/qwen3.5-122b-a10b"},"qwen/qwen3.5-35b-a3b":{"id":"qwen/qwen3.5-35b-a3b","name":"Qwen3.5 35B A3B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.25,"output":1.3},"canonical_model_id":"alibaba/qwen3.5-35b-a3b"},"qwen/qwen3.5-27b":{"id":"qwen/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"reasoning_options":[{"type":"toggle"}],"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"cost":{"input":0.3,"output":2.4},"canonical_model_id":"alibaba/qwen3.5-27b"}}}} \ No newline at end of file